DETAILED ACTION
Notice of AIA Status
The present application is being examined under the AIA the first inventor to file provisions.
Response to Arguments
Applicant’s arguments (see remarks below), filed 02/26/2026, with respect to claims 1, 3, 5, 7-8, 10, 12, 14-15, 18, 20, 22, 24, 26-32 have been fully considered but respectfully are unpersuasive.
On page 8, the applicant argues “With regard to the §103 rejection of claims 1, 3, 8, 10, 15 and 30-32, Applicant respectfully traverses on the ground that the collective teachings of Wang, Fong and Dwivedi fail to disclose each and every limitation of claims 1, 3, 8, 10, 15 and 30-32, arranged as recited in those claims, and on the further ground that there is no teaching, suggestion or motivation in the Wang, Fong and Dwivedi references to modify their collective teachings in a manner that would reach the particular claimed arrangements.”
In response, the Office finds this argument unpersuasive for the reasons stated below.
On page 9, the applicant argues “The Office Action fails to point out where Wang, Fong, Dwivedi, or any of the other cited references, teach or suggest the explicit claim limitations associated with the adapting step. As such, the Office Action fails to present a legally-sufficient, prima facie case of obviousness.”
In response, the Office finds this argument unpersuasive. Based on the breadth of the claim language, the prior art by WANG et al. (US 20110188780 A1) explicitly teaches adapt the selected algorithm to transform a two-dimensional image into a three-dimensional model in the system (Fig. 9. Paragraph [0029]-WANG discloses as shown in FIG. 2, a 2D image is received (step 202). Content of the 2D image may be analyzed (204). The content analysis may include, for example, image categorization, object identification, or the like. Based on a result of the content analysis, a corresponding 2D-3D image conversion method may be chosen or determined (206). Further in paragraph [0064]-WANG discloses based on the historic data, 2D-to-3D image converter 106 may train itself for improved performance);
Although WANG explicitly teaches adapt one or more parameters of the selected algorithm to transform a two-dimensional image into a three-dimensional model in the system.
WANG in view of FONG fails to explicitly teach adapt one or more parameters of the selected algorithm to transform a two-dimensional image into a three-dimensional model in the system based on at least a portion of the one or more accelerator types, the utilization data and the availability data associated with the one or more accelerators of the system.
However, DWIVEDI et al. (US 20210366166 A1) explicitly teaches adapt one or more parameters of the selected algorithm to transform a image into a model in the system (Fig. 2. Paragraph [0023]-DWIVEDI discloses the processors 210s are configured to select reconstructions algorithms to be executed in the optimal weighted execution sequence. The processor 210s may select them from the available reconstruction algorithms based on factors such as: characteristics of available reconstruction algorithms, e.g., processing parameters that the algorithms enhance/reduce, iteration times of the algorithms, points of diminishing return for the algorithms (how effective each reconstruction algorithm is when it is iterated a certain number of times), and the preferred processing parameters (wherein the preferred processing parameters include at least one of a preferred performance parameter such as a preferred amount of time or computing resource for reconstruction, or a preferred visual parameter). Please also see Fig. 3 and read paragraph [0014 and 0035-0038]) based on at least a portion of the one or more accelerator types, the utilization data and the availability data associated with the one or more accelerators of the system (Fig. 2. Paragraph [0026]-DWIVEDI discloses the processors 210s are further configured to determine an optimal weighted execution sequence of the selected reconstruction algorithms. In paragraph [0028]-DWIVEDI discloses the processors 210s then determine a plurality of weighted execution sequences for the representative slice. The weighted execution sequences represent candidates for the optimal weighted execution sequence. The weighted execution sequences are determined based on various factors such as (without limitation) the preferred processing parameter, the characteristics of the selected reconstruction algorithms, the characteristics of the image data, and the time and computing constraints. The time and computing constraints represent the resources available to complete the request, computation time requirement, and available processing power. Please also see Fig. 3 and read paragraph [0023 and 0035-0044]);
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG in view of FONG of having an apparatus, with the teachings of DWIVEDI of having adapt one or more parameters of the selected algorithm to transform a image into a model in the system based on at least a portion of the one or more accelerator types, the utilization data and the availability data associated with the one or more accelerators of the system.
Wherein WANG’s apparatus having adapt one or more parameters of the selected algorithm to transform a two-dimensional image into a three-dimensional model in the system based on at least a portion of the one or more accelerator types, the utilization data and the availability data associated with the one or more accelerators of the system.
The motivation behind the modification would have been to obtain an apparatus that significantly reduces the computational resources, efficiently manages computational load and improves the accuracy, quality and robustness of 3D models, since both WANG and DWIVEDI concern image reconstruction. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images, while DWIVEDI provides systems and methods that can dynamically select the best reconstruction algorithms and sequence of algorithms using significantly smaller amounts of time and computing resources compared to the conventional methods by leveraging parallel processing capabilities of advanced hardware processors and can significantly reduce the time and computing resources spent for necessary pre- and post-processing tasks. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and DWIVEDI et al. (US 20210366166 A1), Abstract and paragraph [0015 and 0057].
On page 9, the applicant argues “Since independent claims 8 and 15 contain limitations similar to those of independent claim 1, claims 8 and 15 are believed to be patentable for at least the reasons described above with respect to claim 1.”
In response, the Office finds this argument unpersuasive for the reasons stated above and below.
On page 9, the applicant argues “Moreover, one or more of the dependent claims are also believed to recite separately patentable subject matter.”
In response, the Office finds this argument unpersuasive for the reasons stated above and below.
On page 9, the applicant argues “The additionally cited Gallo, Wang '054, and Brown references fail to overcome the fundamental deficiencies of Wang, Fong and Dwivedi as applied to the independent claims, and the §103 rejections of the remaining dependent claims are therefore also respectfully traversed.”
In response, the Office finds this argument unpersuasive for the reasons stated above and below.
On page 9, the applicant argues “In view of the above, Applicant respectfully submits that the claims are in condition for allowance, and respectfully requests withdrawal of the rejections.”
In response, the Office finds this argument unpersuasive for the reasons stated above and below.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102 of this title, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1, 3, 8, 10, 15 and 30-32 are rejected under 35 U.S.C. 103 as being unpatentable over WANG et al. (US 20110188780 A1), hereinafter referenced as WANG and in view of FONG et al. (US 20210191759 A1), hereinafter referenced as FONG and in further view of DWIVEDI et al. (US 20210366166 A1), hereinafter referenced as DWIVEDI.
Regarding claim 1, WANG explicitly teaches a method (Fig. 2. Paragraph [0029]-WANG discloses FIG. 2 is a flow chart illustrating an exemplary 2D-to-3D image conversion based on image content. Please also see Fig. 1 and 9 and paragraph [0026 and 0059]), comprising:
detecting an object in a two-dimensional image (Fig. 2. Paragraph [0029]-WANG discloses in FIG. 2, a 2D image is received (step 202) and content of the 2D image may be analyzed (step 204)) and identifying an object type of the detected object (Fig. 2. Paragraph [0029]-WANG discloses the content analysis may include image categorization, object identification, or the like. Further in paragraph [0030]-WANG discloses each object in a 2D image may be identified or classified as one of object categories (or classes). Please also see Fig. 9 and read paragraph [0060-0061]); and
selecting, based on the identified object type of the detected object (Fig. 2. Paragraph [0029]-WANG discloses based on a result of the content analysis, a corresponding 2D-3D image conversion method may be chosen or determined (206). Further in paragraph [0030]-WANG discloses a corresponding method is adopted to convert the 2D image into a 3D image according to the categorization and/or subcategorization), an algorithm from a plurality of algorithms configured to transform a two-dimensional image into a three-dimensional model (Fig. 2. Paragraph [0030]-WANG discloses the conversion method may be but is not limited to, e.g., shifting a left or right eye image, shifting image pixels depending on their positions, shifting edges of objects in the 2D image, shifting image frequencies, creating a disparity between left and right eye images based on a 3D model, creating a 3D image based on a depth map generated based on the result of content analysis, etc. Please also see Fig. 9 and read paragraph [0062]);
wherein the method (Fig. 9. Paragraph [0057]-WANG discloses FIG. 9 is a block diagram illustrating an exemplary 2D-to-3D image converter 106 in the exemplary system 100 of FIG. 1 (wherein the method in Fig. 2 may be performed by exemplary system 100). The 2D-to-3D image converter 106 may include an image content analyzer 902, a conversion method chooser 906, a 3D image generator 908, and an image rendering engine 910) are performed by at least one processor and at least one memory storing executable computer program instructions in the system (Fig. 9. Paragraph [0057]-WANG discloses one or more of the components depicted in FIG. 9 may be implemented in software on one or more computing systems. Such components may comprise one or more software applications, which may comprise one or more computer units including storage devices containing computer-readable instructions which, when executed by a processor, cause a computer to perform steps of a method. Computer-readable instructions may be stored on a tangible non-transitory computer-readable medium).
WANG fails to explicitly teach collecting system resource data comprising data indicative of one or more accelerators within a system including one or more accelerator types, utilization data and availability data associated with the one or more accelerators of the system.
However, FONG explicitly teaches collecting system resource data comprising data indicative of one or more accelerators within a system including one or more accelerator types, utilization data and availability data associated with the one or more accelerators of the system (Fig. 1. Paragraph [0026]-FONG discloses the AI platform (250) including a resource manager (258). Examples of resources include to a quantity of GPUs utilized, GPU types, etc. The profile manager (252) collects data associated with these resources and associated resource consumption from the corresponding ML model, and ML routines that are part of the ML model. In paragraph [0030]-FONG discloses the computing resource allocation may change across the iterations based on the requirements of the ML execution models as well as availability of the resources. In paragraph [0037]-FONG discloses possible execution directives include, setting limits of memory usage, and allocating resources for utilization. With respect to memory usage, limits may be set for application instances based on one or more previous iterations or historical data from one or more previous runs).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG of having a method comprising: detecting an object in a two-dimensional image and identifying an object type of the detected object, with the teachings of FONG of having collecting system resource data comprising data indicative of one or more accelerators within a system including one or more accelerator types, utilization data and availability data associated with the one or more accelerators of the system.
Wherein WANG’s method having collecting system resource data comprising data indicative of one or more accelerators within a system including one or more accelerator types, utilization data and availability data associated with the one or more accelerators of the system.
The motivation behind the modification would have been to obtain a method that improve the functionality and operation of an artificial intelligence platform to model ML, resource usage, and the accuracy, quality and robustness of 3D models, since both WANG and FONG concern machine learning. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images using a learning model, while FONG provides systems and methods that improve the functionality and operation of an artificial intelligence platform to model ML application execution and resource usage, including efficient and effective ML application performance and resource management. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and FONG et al. (US 20210191759 A1), Abstract and paragraph [0077].
Although WANG explicitly teaches adapting the selected algorithm to transform a two-dimensional image into a three-dimensional model in the system (Fig. 9. Paragraph [0029]-WANG discloses as shown in FIG. 2, a 2D image is received (step 202). Content of the 2D image may be analyzed (204). The content analysis may include, for example, image categorization, object identification, or the like. Based on a result of the content analysis, a corresponding 2D-3D image conversion method may be chosen or determined (206). Further in paragraph [0064]-WANG discloses based on the historic data, 2D-to-3D image converter 106 may train itself for improved performance);
testing the adapted algorithm to determine if the adapted algorithm will transform the two-dimensional image into the three-dimensional model in the system including utilization data of the system (Fig. 9. Paragraph [0064]-WANG discloses during the above-described 2D-to-3D image conversion based on image content, each component of 2D-to-3D image converter 106 may store its computation/determination results in image database 904 for later retrieval or training purpose. Based on the historic data, 2D-to-3D image converter 106 may train itself for improved performance. Please also read paragraph [0066] (wherein methods may be implemented by an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), a digital signal processor (DSP));
executing the adapted algorithm to transform the two-dimensional image into the three-dimensional model in the system (Fig. 9. Paragraph [0029]-WANG discloses applying the chosen method to the 2D image, a 3D image can be generated (step 208));
WANG in view of FONG fails to explicitly teach adapting one or more parameters of the selected algorithm to transform a two-dimensional image into a three-dimensional model in the system based on at least a portion of the one or more accelerator types, the utilization data and the availability data associated with the one or more accelerators of the system; using at least one of the one or more accelerators of the system; in response to the testing being successful, executing the adapted algorithm to transform the two-dimensional image into the three-dimensional model in the system.
However, DWIVEDI explicitly teaches adapting one or more parameters of the selected algorithm to transform a two-dimensional image into a three-dimensional model in the system (Fig. 2. Paragraph [0023]-DWIVEDI discloses the processors 210s are configured to select reconstructions algorithms to be executed in the optimal weighted execution sequence. The processor 210s may select them from the available reconstruction algorithms based on factors such as: characteristics of available reconstruction algorithms, e.g., processing parameters that the algorithms enhance/reduce, iteration times of the algorithms, points of diminishing return for the algorithms (how effective each reconstruction algorithm is when it is iterated a certain number of times), and the preferred processing parameters (wherein the preferred processing parameters include at least one of a preferred performance parameter such as a preferred amount of time or computing resource for reconstruction, or a preferred visual parameter). Please also see Fig. 3 and read paragraph [0014 and 0035-0038]) based on at least a portion of the one or more accelerator types, the utilization data and the availability data associated with the one or more accelerators of the system (Fig. 2. Paragraph [0026]-DWIVEDI discloses the processors 210s are further configured to determine an optimal weighted execution sequence of the selected reconstruction algorithms. In paragraph [0028]-DWIVEDI discloses the processors 210s then determine a plurality of weighted execution sequences for the representative slice. The weighted execution sequences represent candidates for the optimal weighted execution sequence. The weighted execution sequences are determined based on various factors such as (without limitation) the preferred processing parameter, the characteristics of the selected reconstruction algorithms, the characteristics of the image data, and the time and computing constraints. The time and computing constraints represent the resources available to complete the request, computation time requirement, and available processing power. Please also see Fig. 3 and read paragraph [0023 and 0035-0044]);
testing the adapted algorithm (Fig. 2. Paragraph [0015]-DWIVEDI discloses the technique dynamically determines the optimal weighted sequence for two or more algorithms by testing out multiple weighted sequences during reconstruction and choosing the best sequence. The technique tests multiple weighted sequences in parallel using the parallel processing capability of the multi-processor unit) to determine if the adapted algorithm will transform the image into the model in the system (Fig. 4. Paragraph [0055]-DWIVEDI discloses first image 410 depicts an image reconstructed from executing a first algorithm (algorithm A) for 200 iterations (wherein reconstruction took approximately 130 seconds and the result indicates a lack of detail but suitability for removing noise). In paragraph [0056]-DWIVEDI discloses second image 420 depicts an image reconstructed from executing a second algorithm (algorithm B) for 200 iterations (wherein reconstruction took approximately 60 seconds and indicates a suitability for retaining the details, but not removal of noise). In paragraph [0057]-DWIVEDI discloses third image 430 depicts an image reconstructed from executing the first algorithm for the first 100 iterations and the second algorithm for the next 100 iterations (wherein the reconstruction took 75 seconds and indicates a relatively high level of detail and partial removal of noises)) using at least one of the one or more accelerators (Fig. 2, #110 and 210-N, called multi-processor unit and a processor, respectively. Paragraph [0017 and 0019]. Further in paragraph [0017]-DWIVEDI discloses the multi-processor unit 110 is configured to determine an optimal weighted execution sequence of reconstruction algorithms for processing image data from one or more of the devices. The multi-processor unit 110 may be any processing unit that is capable of parallel-processing, such as a graphics processing unit (GPU), a central processing unit (CPU), or a combination of both CPU and GPU) of the system (Fig. 1, #100 called a system. Paragraph [0016]-DWIVEDI discloses FIG. 1 illustrates an embodiment of a system 100 for visualizing image data);
in response to the testing being successful, executing the adapted algorithm to transform the image into the model in the system (Fig. 2. Paragraph [0048]-DWIVEDI discloses based on the comparison of the reconstructed images, one of the weighted sequences that provides the best figure of merit (FOM) for the preferred processing parameter is selected as the optimal weighted execution sequence at step 360. In paragraph [0052]-DWIVEDI discloses at step 370, the selected algorithms are executed according to the optimal weighted execution sequence. Each of the selected reconstruction algorithms is executed on the image data iteratively based on the respective weight, and also sequentially based on the execution order).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG in view of FONG of having a method comprising: detecting an object in a two-dimensional image and identifying an object type of the detected object, with the teachings of DWIVEDI of having adapting one or more parameters of the selected algorithm to transform a two-dimensional image into a three-dimensional model in the system based on at least a portion of the one or more accelerator types, the utilization data and the availability data associated with the one or more accelerators of the system; testing the adapted algorithm to determine if the adapted algorithm will transform the image into the model in the system using at least one of the one or more accelerators in response to the testing being successful, executing the adapted algorithm to transform the image into the model in the system.
Wherein WANG’s method having adapting one or more parameters of the selected algorithm to transform a two-dimensional image into a three-dimensional model in the system based on at least a portion of the one or more accelerator types, the utilization data and the availability data associated with the one or more accelerators of the system; testing the adapted algorithm to determine if the adapted algorithm will transform the two- dimensional image into the three-dimensional model in the system using at least one of the one or more accelerators of the system; and in response to the testing being successful, executing the adapted algorithm to transform the two-dimensional image into the three-dimensional model in the system.
The motivation behind the modification would have been to obtain a method that significantly reduces the computational resources, efficiently manages computational load and improves the accuracy, quality and robustness of 3D models, since both WANG and DWIVEDI concern image reconstruction. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images, while DWIVEDI provides systems and methods that can dynamically select the best reconstruction algorithms and sequence of algorithms using significantly smaller amounts of time and computing resources compared to the conventional methods by leveraging parallel processing capabilities of advanced hardware processors and can significantly reduce the time and computing resources spent for necessary pre- and post-processing tasks. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and DWIVEDI et al. (US 20210366166 A1), Abstract and paragraph [0015 and 0057].
Regarding claim 3, WANG in view of FONG and in further view of DWIVEDI explicitly teaches the method of claim 1, WANG fails to explicitly teach wherein the collected system resource data further comprises data indicative of versions and quantities of the one or more accelerators within the system, past utilization data, and current availability of the one or more accelerators within the system.
However, FONG explicitly teaches wherein the collected system resource data further comprises data indicative of versions and quantities of the one or more accelerators within the system, past utilization data, and current availability of the one or more accelerators within the system (Fig. 1. Paragraph [0026]-FONG discloses the AI platform (250) including a resource manager (258). Examples of resources include to a quantity of GPUs utilized, GPU types, etc. The profile manager (252) collects data associated with these resources and associated resource consumption from the corresponding ML model, and ML routines that are part of the ML model. In paragraph [0030]-FONG discloses the computing resource allocation may change across the iterations based on the requirements of the ML execution models as well as availability of the resources. In paragraph [0037]-FONG discloses possible execution directives include, setting limits of memory usage, and allocating resources for utilization. With respect to memory usage, limits may be set for application instances based on one or more previous iterations or historical data from one or more previous runs).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG in view of FONG and in further view of DWIVEDI of having a method comprising: detecting an object in a two-dimensional image and identifying an object type of the detected object, with the teachings of FONG of having wherein the collected system resource data further comprises data indicative of versions and quantities of the one or more accelerators within the system, past utilization data, and current availability of the one or more accelerators within the system.
Wherein WANG’s method having wherein the collected system resource data further comprises data indicative of versions and quantities of the one or more accelerators within the system, past utilization data, and current availability of the one or more accelerators within the system.
The motivation behind the modification would have been to obtain a method that improve the functionality and operation of an artificial intelligence platform to model ML, resource usage, and the accuracy, quality and robustness of 3D models, since both WANG and FONG concern machine learning. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images using a learning model, while FONG provides systems and methods that improve the functionality and operation of an artificial intelligence platform to model ML application execution and resource usage, including efficient and effective ML application performance and resource management. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and FONG et al. (US 20210191759 A1), Abstract and paragraph [0077].
Regarding claim 8, WANG teaches an apparatus (Fig. 1, #100 called a system. Paragraph [0020]), comprising:
at least one processor and at least one memory storing computer program instructions wherein, when the at least one processor executes the computer program instructions (Fig. 1. Paragraph [0026]-WANG discloses 2D-to-3D image converter 106 can be implemented as a software program executing in a processor and/or as hardware that performs a 2D-to-3D image conversion based on image content. Please also read paragraph [0066]), the apparatus is configured to:
detect an object in a two-dimensional image (Fig. 2. Paragraph [0029]-WANG discloses in FIG. 2, a 2D image is received (step 202) and content of the 2D image may be analyzed (step 204)) and identifying an object type of the detected object (Fig. 2. Paragraph [0029]-WANG discloses the content analysis may include image categorization, object identification, or the like. Further in paragraph [0030]-WANG discloses each object in a 2D image may be identified or classified as one of object categories (or classes). Please also see Fig. 9 and read paragraph [0060-0061]);
select, based on the identified object type of the detected object (Fig. 2. Paragraph [0029]-WANG discloses based on a result of the content analysis, a corresponding 2D-3D image conversion method may be chosen or determined (206). Further in paragraph [0030]-WANG discloses a corresponding method is adopted to convert the 2D image into a 3D image according to the categorization and/or subcategorization), an algorithm from a plurality of algorithms configured to transform a two-dimensional image into a three-dimensional model (Fig. 2. Paragraph [0030]-WANG discloses the conversion method may be but is not limited to, e.g., shifting a left or right eye image, shifting image pixels depending on their positions, shifting edges of objects in the 2D image, shifting image frequencies, creating a disparity between left and right eye images based on a 3D model, creating a 3D image based on a depth map generated based on the result of content analysis, etc. Please also see Fig. 9 and read paragraph [0062]);
WANG fails to explicitly teach collect system resource data comprising data indicative of one or more accelerators within a system including one or more accelerator types, utilization data and availability data associated with the one or more accelerators of the system;
However, FONG explicitly teaches collect system resource data comprising data indicative of one or more accelerators within a system including one or more accelerator types, utilization data and availability data associated with the one or more accelerators of the system (Fig. 1. Paragraph [0026]-FONG discloses the AI platform (250) including a resource manager (258). Examples of resources include to a quantity of GPUs utilized, GPU types, etc. The profile manager (252) collects data associated with these resources and associated resource consumption from the corresponding ML model, and ML routines that are part of the ML model. In paragraph [0030]-FONG discloses the computing resource allocation may change across the iterations based on the requirements of the ML execution models as well as availability of the resources. In paragraph [0037]-FONG discloses possible execution directives include, setting limits of memory usage, and allocating resources for utilization. With respect to memory usage, limits may be set for application instances based on one or more previous iterations or historical data from one or more previous runs).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG of having an apparatus comprising: detecting an object in a two-dimensional image and identifying an object type of the detected object, with the teachings of FONG of having collect system resource data comprising data indicative of one or more accelerators within a system including one or more accelerator types, utilization data and availability data associated with the one or more accelerators of the system.
Wherein WANG’s apparatus having collect system resource data comprising data indicative of one or more accelerators within a system including one or more accelerator types, utilization data and availability data associated with the one or more accelerators of the system.
The motivation behind the modification would have been to obtain an apparatus that improve the functionality and operation of an artificial intelligence platform to model ML, resource usage, and the accuracy, quality and robustness of 3D models, since both WANG and FONG concern machine learning. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images using a learning model, while FONG provides systems and methods that improve the functionality and operation of an artificial intelligence platform to model ML application execution and resource usage, including efficient and effective ML application performance and resource management. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and FONG et al. (US 20210191759 A1), Abstract and paragraph [0077].
Although WANG explicitly teaches adapt the selected algorithm to transform a two-dimensional image into a three-dimensional model in the system (Fig. 9. Paragraph [0029]-WANG discloses as shown in FIG. 2, a 2D image is received (step 202). Content of the 2D image may be analyzed (204). The content analysis may include, for example, image categorization, object identification, or the like. Based on a result of the content analysis, a corresponding 2D-3D image conversion method may be chosen or determined (206). Further in paragraph [0064]-WANG discloses based on the historic data, 2D-to-3D image converter 106 may train itself for improved performance);
test the adapted algorithm to determine if the adapted algorithm will transform the two- dimensional image into the three-dimensional model in the system (Fig. 9. Paragraph [0064]-WANG discloses during the above-described 2D-to-3D image conversion based on image content, each component of 2D-to-3D image converter 106 may store its computation/determination results in image database 904 for later retrieval or training purpose. Based on the historic data, 2D-to-3D image converter 106 may train itself for improved performance. Please also read paragraph [0066] (wherein methods may be implemented by an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), a digital signal processor (DSP));
executing the adapted algorithm to transform the two-dimensional image into the three-dimensional model in the system (Fig. 9. Paragraph [0029]-WANG discloses applying the chosen method to the 2D image, a 3D image can be generated (step 208)).
WANG in view of FONG fails to explicitly teach adapt one or more parameters of the selected algorithm to transform a two-dimensional image into a three-dimensional model in the system based on at least a portion of the one or more accelerator types, the utilization data and the availability data associated with the one or more accelerators of the system; using at least one of the one or more accelerators of the system; in response to the testing being successful, executing the adapted algorithm to transform the two-dimensional image into the three-dimensional model in the system.
However, DWIVEDI explicitly teaches adapt one or more parameters of the selected algorithm to transform a image into a model in the system (Fig. 2. Paragraph [0023]-DWIVEDI discloses the processors 210s are configured to select reconstructions algorithms to be executed in the optimal weighted execution sequence. The processor 210s may select them from the available reconstruction algorithms based on factors such as: characteristics of available reconstruction algorithms, e.g., processing parameters that the algorithms enhance/reduce, iteration times of the algorithms, points of diminishing return for the algorithms (how effective each reconstruction algorithm is when it is iterated a certain number of times), and the preferred processing parameters (wherein the preferred processing parameters include at least one of a preferred performance parameter such as a preferred amount of time or computing resource for reconstruction, or a preferred visual parameter). Please also see Fig. 3 and read paragraph [0014 and 0035-0038]) based on at least a portion of the one or more accelerator types, the utilization data and the availability data associated with the one or more accelerators of the system (Fig. 2. Paragraph [0026]-DWIVEDI discloses the processors 210s are further configured to determine an optimal weighted execution sequence of the selected reconstruction algorithms. In paragraph [0028]-DWIVEDI discloses the processors 210s then determine a plurality of weighted execution sequences for the representative slice. The weighted execution sequences represent candidates for the optimal weighted execution sequence. The weighted execution sequences are determined based on various factors such as (without limitation) the preferred processing parameter, the characteristics of the selected reconstruction algorithms, the characteristics of the image data, and the time and computing constraints. The time and computing constraints represent the resources available to complete the request, computation time requirement, and available processing power. Please also see Fig. 3 and read paragraph [0023 and 0035-0044]);
test the adapted algorithm (Fig. 2. Paragraph [0015]-DWIVEDI discloses the technique dynamically determines the optimal weighted sequence for two or more algorithms by testing out multiple weighted sequences during reconstruction and choosing the best sequence. The technique tests multiple weighted sequences in parallel using the parallel processing capability of the multi-processor unit) to determine if the adapted algorithm will transform the image into the model in the system (Fig. 4. Paragraph [0055]-DWIVEDI discloses first image 410 depicts an image reconstructed from executing a first algorithm (algorithm A) for 200 iterations (wherein reconstruction took approximately 130 seconds and the result indicates a lack of detail but suitability for removing noise). In paragraph [0056]-DWIVEDI discloses second image 420 depicts an image reconstructed from executing a second algorithm (algorithm B) for 200 iterations (wherein reconstruction took approximately 60 seconds and indicates a suitability for retaining the details, but not removal of noise). In paragraph [0057]-DWIVEDI discloses third image 430 depicts an image reconstructed from executing the first algorithm for the first 100 iterations and the second algorithm for the next 100 iterations (wherein the reconstruction took 75 seconds and indicates a relatively high level of detail and partial removal of noises)) using at least one of the one or more accelerators (Fig. 2, #110 and 210-N, called multi-processor unit and a processor, respectively. Paragraph [0017 and 0019]. Further in paragraph [0017]-DWIVEDI discloses the multi-processor unit 110 is configured to determine an optimal weighted execution sequence of reconstruction algorithms for processing image data from one or more of the devices. The multi-processor unit 110 may be any processing unit that is capable of parallel-processing, such as a graphics processing unit (GPU), or a combination of both CPU and GPU) of the system (Fig. 1, #100 called a system. Paragraph [0016]-DWIVEDI discloses FIG. 1 illustrates an embodiment of a system 100 for visualizing image data).
in response to the testing being successful, executing the adapted algorithm to transform the image into the model in the system (Fig. 2. Paragraph [0048]-DWIVEDI discloses based on the comparison of the reconstructed images, one of the weighted sequences that provides the best figure of merit (FOM) for the preferred processing parameter is selected as the optimal weighted execution sequence at step 360. In paragraph [0052]-DWIVEDI discloses at step 370, the selected algorithms are executed according to the optimal weighted execution sequence. Each of the selected reconstruction algorithms is executed on the image data iteratively based on the respective weight, and also sequentially based on the execution order).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of in view of FONG of having an apparatus comprising: detecting an object in a two-dimensional image and identifying an object type of the detected object, with the teachings of DWIVEDI of having adapt one or more parameters of the selected algorithm to transform a image into a model in the system based on at least a portion of the one or more accelerator types, the utilization data and the availability data associated with the one or more accelerators of the system; test the adapted algorithm to determine if the adapted algorithm will transform the image into the model in the system using at least one of the one or more accelerators of the system; in response to the testing being successful, executing the adapted algorithm to transform the image into the model in the system.
Wherein WANG’s apparatus having adapt one or more parameters of the selected algorithm to transform a two-dimensional image into a three-dimensional model in the system based on at least a portion of the one or more accelerator types, the utilization data and the availability data associated with the one or more accelerators of the system; test the adapted algorithm to determine if the adapted algorithm will transform the two-dimensional image into the three-dimensional model in the system using at least one of the one or more accelerators of the system; in response to the testing being successful, executing the adapted algorithm to transform the two-dimensional image into the three-dimensional model in the system.
The motivation behind the modification would have been to obtain an apparatus that significantly reduces the computational resources, efficiently manages computational load and improves the accuracy, quality and robustness of 3D models, since both WANG and DWIVEDI concern image reconstruction. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images, while DWIVEDI provides systems and methods that can dynamically select the best reconstruction algorithms and sequence of algorithms using significantly smaller amounts of time and computing resources compared to the conventional methods by leveraging parallel processing capabilities of advanced hardware processors and can significantly reduce the time and computing resources spent for necessary pre- and post-processing tasks. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and DWIVEDI et al. (US 20210366166 A1), Abstract and paragraph [0015 and 0057].
Regarding claim 10, WANG in view of FONG and in further view of DWIVEDI explicitly teaches the apparatus of claim 8, WANG fails to explicitly teach wherein the collected system resource data further comprises data indicative of versions and quantities of the one or more accelerators within the system, past utilization data, and current availability of the one or more accelerators within the system.
However, FONG explicitly teaches wherein the collected system resource data further comprises data indicative of versions and quantities of the one or more accelerators within the system, past utilization data, and current availability of the one or more accelerators within the system (Fig. 1. Paragraph [0026]-FONG discloses the AI platform (250) including a resource manager (258). Examples of resources include to a quantity of GPUs utilized, GPU types, etc. The profile manager (252) collects data associated with these resources and associated resource consumption from the corresponding ML model, and ML routines that are part of the ML model. In paragraph [0030]-FONG discloses the computing resource allocation may change across the iterations based on the requirements of the ML execution models as well as availability of the resources. In paragraph [0037]-FONG discloses possible execution directives include, setting limits of memory usage, and allocating resources for utilization. With respect to memory usage, limits may be set for application instances based on one or more previous iterations or historical data from one or more previous runs).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG in view of FONG and in further view of DWIVEDI of having an apparatus comprising: detecting an object in a two-dimensional image and identifying an object type of the detected object, with the teachings of FONG of having wherein the collected system resource data further comprises data indicative of versions and quantities of the one or more accelerators within the system, past utilization data, and current availability of the one or more accelerators within the system.
Wherein WANG’s apparatus having wherein the collected system resource data further comprises data indicative of versions and quantities of the one or more accelerators within the system, past utilization data, and current availability of the one or more accelerators within the system.
The motivation behind the modification would have been to obtain an apparatus that improve the functionality and operation of an artificial intelligence platform to model ML, resource usage, and the accuracy, quality and robustness of 3D models, since both WANG and FONG concern machine learning. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images using a learning model, while FONG provides systems and methods that improve the functionality and operation of an artificial intelligence platform to model ML application execution and resource usage, including efficient and effective ML application performance and resource management. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and FONG et al. (US 20210191759 A1), Abstract and paragraph [0077].
Regarding claim 15, WANG explicitly teaches a computer program product (Fig. 9, #106 called an exemplary 2D-to-3D image converter. Paragraph [0057]. Further in paragraph [0065]-WANG discloses the methods may be implemented as a computer program product. Please also see Fig. 1 and read paragraph [0026]) stored on a non-transitory computer-readable medium and comprising machine executable instructions, the machine executable instructions, when executed, causing a processing device to perform steps (Fig. 9. Paragraph [0057]-WANG discloses FIG. 9 is a block diagram illustrating an exemplary 2D-to-3D image converter 106 in the exemplary system 100 of FIG. 1. Such components may comprise one or more software applications, which may comprise one or more computer units including storage devices containing computer-readable instructions which, when executed by a processor, cause a computer to perform steps of a method (wherein the exemplary 2D-to-3D image converter 106 may implement method 2). Computer-readable instructions may be stored on a tangible non-transitory computer-readable medium) of:
selecting, based on the identified object type of the detected object (Fig. 2. Paragraph [0029]-WANG discloses in FIG. 2, a 2D image is received (step 202) and content of the 2D image may be analyzed (step 204). The content analysis may include image categorization, object identification, or the like. Further in paragraph [0030]-WANG discloses each object in a 2D image may be identified or classified as one of object categories (or classes)), an algorithm (Fig. 2. Paragraph [0030]-WANG discloses based on a result of the content analysis, a corresponding 2D-3D image conversion method may be chosen or determined (206)) from a plurality of algorithms configured to transform a two-dimensional image into a three-dimensional model (Fig. 2. Paragraph [0030]-WANG discloses the conversion method may be shifting a left or right eye image, shifting image pixels depending on their positions, shifting edges of objects in the 2D image, shifting image frequencies, creating a disparity between left and right eye images based on a 3D model, creating a 3D image based on a depth map generated based on the result of content analysis, etc. Please also see Fig. 9 and read paragraph [0062]);
WANG fails to explicitly teach collecting system resource data comprising data indicative of one or more accelerators within a system including one or more accelerator types, utilization data and availability data associated with the one or more accelerators of the system;
However, FONG explicitly teaches collecting system resource data comprising data indicative of one or more accelerators within a system including one or more accelerator types, utilization data and availability data associated with the one or more accelerators of the system (Fig. 1. Paragraph [0026]-FONG discloses the AI platform (250) including a resource manager (258). Examples of resources include to a quantity of GPUs utilized, GPU types, etc. The profile manager (252) collects data associated with these resources and associated resource consumption from the corresponding ML model, and ML routines that are part of the ML model. In paragraph [0030]-FONG discloses the computing resource allocation may change across the iterations based on the requirements of the ML execution models as well as availability of the resources. In paragraph [0037]-FONG discloses possible execution directives include, setting limits of memory usage, and allocating resources for utilization. With respect to memory usage, limits may be set for application instances based on one or more previous iterations or historical data from one or more previous runs).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG of having a computer program product stored on a non-transitory computer-readable medium and comprising machine executable instructions, the machine executable instructions, when executed, causing a processing device to perform steps, with the teachings of FONG of having collecting system resource data comprising data indicative of one or more accelerators within a system including one or more accelerator types, utilization data and availability data associated with the one or more accelerators of the system.
Wherein WANG’s computer program product having collecting system resource data comprising data indicative of one or more accelerators within a system including one or more accelerator types, utilization data and availability data associated with the one or more accelerators of the system.
The motivation behind the modification would have been to obtain a computer program product that improves the functionality and operation of an artificial intelligence platform to model ML, resource usage, and the accuracy, quality and robustness of 3D models, since both WANG and FONG concern machine learning. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images using a learning model, while FONG provides systems and methods that improve the functionality and operation of an artificial intelligence platform to model ML application execution and resource usage, including efficient and effective ML application performance and resource management. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and FONG et al. (US 20210191759 A1), Abstract and paragraph [0077].
Although WANG explicitly teaches adapting the selected algorithm to transform a two-dimensional image into a three-dimensional model in the system (Fig. 9. Paragraph [0029]-WANG discloses as shown in FIG. 2, a 2D image is received (step 202). Content of the 2D image may be analyzed (204). The content analysis may include, for example, image categorization, object identification, or the like. Based on a result of the content analysis, a corresponding 2D-3D image conversion method may be chosen or determined (206). Further in paragraph [0064]-WANG discloses based on the historic data, 2D-to-3D image converter 106 may train itself for improved performance);
testing the adapted algorithm to determine if the adapted algorithm will transform the two- dimensional image into the three-dimensional model in the system (Fig. 9. Paragraph [0064]-WANG discloses during the above-described 2D-to-3D image conversion based on image content, each component of 2D-to-3D image converter 106 may store its computation/determination results in image database 904 for later retrieval or training purpose. Based on the historic data, 2D-to-3D image converter 106 may train itself for improved performance. Please also read paragraph [0066] (wherein methods may be implemented by an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), a digital signal processor (DSP)); and
executing the adapted algorithm to transform the two-dimensional image into the three-dimensional model in the system (Fig. 9. Paragraph [0029]-WANG discloses applying the chosen method to the 2D image, a 3D image can be generated (step 208).
WANG fails to explicitly teach adapting one or more parameters of the selected algorithm to transform a two-dimensional image into a three-dimensional model in the system based on at least a portion of the one or more accelerator types, the utilization data and the availability data associated with the one or more accelerators of the system; using at least one of the one or more accelerators of the system; in response to the testing being successful, executing the adapted algorithm to transform the two-dimensional image into the three-dimensional model in the system.
However, DWIVEDI explicitly teaches adapting one or more parameters of the selected algorithm to transform a image into a model in the system (Fig. 2. Paragraph [0023]-DWIVEDI discloses the processors 210s are configured to select reconstructions algorithms to be executed in the optimal weighted execution sequence. The processor 210s may select them from the available reconstruction algorithms based on factors such as: characteristics of available reconstruction algorithms, e.g., processing parameters that the algorithms enhance/reduce, iteration times of the algorithms, points of diminishing return for the algorithms (how effective each reconstruction algorithm is when it is iterated a certain number of times), and the preferred processing parameters (wherein the preferred processing parameters include at least one of a preferred performance parameter such as a preferred amount of time or computing resource for reconstruction, or a preferred visual parameter). Please also see Fig. 3 and read paragraph [0014 and 0035-0038]) based on at least a portion of the one or more accelerator types, the utilization data and the availability data associated with the one or more accelerators of the system (Fig. 2. Paragraph [0026]-DWIVEDI discloses the processors 210s are further configured to determine an optimal weighted execution sequence of the selected reconstruction algorithms. In paragraph [0028]-DWIVEDI discloses the processors 210s then determine a plurality of weighted execution sequences for the representative slice. The weighted execution sequences represent candidates for the optimal weighted execution sequence. The weighted execution sequences are determined based on various factors such as (without limitation) the preferred processing parameter, the characteristics of the selected reconstruction algorithms, the characteristics of the image data, and the time and computing constraints. The time and computing constraints represent the resources available to complete the request, computation time requirement, and available processing power. Please also see Fig. 3 and read paragraph [0023 and 0035-0044]);
testing the adapted algorithm (Fig. 2. Paragraph [0015]-DWIVEDI discloses the technique dynamically determines the optimal weighted sequence for two or more algorithms by testing out multiple weighted sequences during reconstruction and choosing the best sequence. The technique tests multiple weighted sequences in parallel using the parallel processing capability of the multi-processor unit) to determine if the adapted algorithm will transform the image into the model in the system (Fig. 4. Paragraph [0055]-DWIVEDI discloses first image 410 depicts an image reconstructed from executing a first algorithm (algorithm A) for 200 iterations (wherein reconstruction took approximately 130 seconds and the result indicates a lack of detail but suitability for removing noise). In paragraph [0056]-DWIVEDI discloses second image 420 depicts an image reconstructed from executing a second algorithm (algorithm B) for 200 iterations (wherein reconstruction took approximately 60 seconds and indicates a suitability for retaining the details, but not removal of noise). In paragraph [0057]-DWIVEDI discloses third image 430 depicts an image reconstructed from executing the first algorithm for the first 100 iterations and the second algorithm for the next 100 iterations (wherein the reconstruction took 75 seconds and indicates a relatively high level of detail and partial removal of noises)) using at least one of the one or more accelerators (Fig. 2, #110 and 210-N, called multi-processor unit and a processor, respectively. Paragraph [0017 and 0019]. Further in paragraph [0017]-DWIVEDI discloses the multi-processor unit 110 is configured to determine an optimal weighted execution sequence of reconstruction algorithms for processing image data from one or more of the devices. The multi-processor unit 110 may be any processing unit that is capable of parallel-processing, such as a graphics processing unit (GPU), or a combination of both CPU and GPU) of the system (Fig. 1, #100 called a system. Paragraph [0016]-DWIVEDI discloses FIG. 1 illustrates an embodiment of a system 100 for visualizing image data); and
in response to the testing being successful, executing the adapted algorithm to transform the image into the model in the system (Fig. 2. Paragraph [0048]-DWIVEDI discloses based on the comparison of the reconstructed images, one of the weighted sequences that provides the best figure of merit (FOM) for the preferred processing parameter is selected as the optimal weighted execution sequence at step 360. In paragraph [0052]-DWIVEDI discloses at step 370, the selected algorithms are executed according to the optimal weighted execution sequence. Each of the selected reconstruction algorithms is executed on the image data iteratively based on the respective weight, and also sequentially based on the execution order).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG in view of FONG of having a computer program product stored on a non-transitory computer-readable medium and comprising machine executable instructions, the machine executable instructions, when executed, causing a processing device to perform steps, with the teachings of DWIVEDI of having adapting one or more parameters of the selected algorithm to transform a image into model in the system based on at least a portion of the one or more accelerator types, the utilization data and the availability data associated with the one or more accelerators of the system; testing the adapted algorithm to determine if the adapted algorithm will transform the image into the model in the system using at least one of the one or more accelerators of the system; and in response to the testing being successful, executing the adapted algorithm to transform the image into the model in the system.
Wherein WANG’s computer program product having adapting one or more parameters of the selected algorithm to transform a two-dimensional image into a three-dimensional model in the system based on at least a portion of the one or more accelerator types, the utilization data and the availability data associated with the one or more accelerators of the system; testing the adapted algorithm to determine if the adapted algorithm will transform the two-dimensional image into the three-dimensional model in the system using at least one of the one or more accelerators of the system; and in response to the testing being successful, executing the adapted algorithm to transform the two-dimensional image into the three-dimensional model in the system.
The motivation behind the modification would have been to obtain a computer program product that significantly reduces the computational resources, efficiently manages computational load and improves the accuracy, quality and robustness of 3D models, since both WANG and DWIVEDI concern image reconstruction. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images, while DWIVEDI provides systems and methods that can dynamically select the best reconstruction algorithms and sequence of algorithms using significantly smaller amounts of time and computing resources compared to the conventional methods by leveraging parallel processing capabilities of advanced hardware processors and can significantly reduce the time and computing resources spent for necessary pre- and post-processing tasks. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and DWIVEDI et al. (US 20210366166 A1), Abstract and paragraph [0015 and 0057].
Regarding claim 30, WANG in view of FONG and in further view of DWIVEDI explicitly teach the method of claim 1, WANG fails to explicitly teach wherein the collected system resource data further comprises data indicative of the availability of one or more graphics processing units within the system.
However, FONG explicitly teaches wherein the collected system resource data further comprises data indicative of the availability of one or more graphics processing units within the system (Fig. 1. Paragraph [0026]-FONG discloses the AI platform (250) including a resource manager (258). Examples of resources include to a quantity of GPUs utilized, GPU types, etc. The profile manager (252) collects data associated with these resources and associated resource consumption from the corresponding ML model, and ML routines that are part of the ML model. In paragraph [0030]-FONG discloses the computing resource allocation may change across the iterations based on the requirements of the ML execution models as well as availability of the resources. In paragraph [0037]-FONG discloses possible execution directives include, setting limits of memory usage, and allocating resources for utilization. With respect to memory usage, limits may be set for application instances based on one or more previous iterations or historical data from one or more previous runs).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG in view of FONG and in further view of DWIVEDI of having a method comprising: detecting an object in a two-dimensional image and identifying an object type of the detected object, with the teachings of FONG of having wherein the collected system resource data further comprises data indicative of the availability of one or more graphics processing units within the system.
Wherein WANG’s method having wherein the collected system resource data further comprises data indicative of the availability of one or more graphics processing units within the system.
The motivation behind the modification would have been to obtain a method that improve the functionality and operation of an artificial intelligence platform to model ML, resource usage, and the accuracy, quality and robustness of 3D models, since both WANG and FONG concern machine learning. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images using a learning model, while FONG provides systems and methods that improve the functionality and operation of an artificial intelligence platform to model ML application execution and resource usage, including efficient and effective ML application performance and resource management. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and FONG et al. (US 20210191759 A1), Abstract and paragraph [0077].
Regarding claim 31, WANG in view of FONG and in further view of DWIVEDI explicitly teach the apparatus of claim 8, WANG fails to explicitly teach wherein the collected system resource data further comprises data indicative of the availability of one or more graphics processing units within the system.
However, FONG explicitly teaches wherein the collected system resource data further comprises data indicative of the availability of one or more graphics processing units within the system (Fig. 1. Paragraph [0026]-FONG discloses the AI platform (250) including a resource manager (258). Examples of resources include to a quantity of GPUs utilized, GPU types, etc. The profile manager (252) collects data associated with these resources and associated resource consumption from the corresponding ML model, and ML routines that are part of the ML model. In paragraph [0030]-FONG discloses the computing resource allocation may change across the iterations based on the requirements of the ML execution models as well as availability of the resources. In paragraph [0037]-FONG discloses possible execution directives include, setting limits of memory usage, and allocating resources for utilization. With respect to memory usage, limits may be set for application instances based on one or more previous iterations or historical data from one or more previous runs).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG in view of FONG and in further view of DWIVEDI of having an apparatus comprising: detecting an object in a two-dimensional image and identifying an object type of the detected object, with the teachings of FONG of having wherein the collected system resource data further comprises data indicative of the availability of one or more graphics processing units within the system.
Wherein WANG’s apparatus having wherein the collected system resource data further comprises data indicative of the availability of one or more graphics processing units within the system.
The motivation behind the modification would have been to obtain an apparatus that improve the functionality and operation of an artificial intelligence platform to model ML, resource usage, and the accuracy, quality and robustness of 3D models, since both WANG and FONG concern machine learning. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images using a learning model, while FONG provides systems and methods that improve the functionality and operation of an artificial intelligence platform to model ML application execution and resource usage, including efficient and effective ML application performance and resource management. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and FONG et al. (US 20210191759 A1), Abstract and paragraph [0077].
Regarding claim 32, WANG in view of FONG and in further view of DWIVEDI explicitly teach the computer program product of claim 15, WANG fails to explicitly teach wherein the collected system resource data further comprises data indicative of the availability of one or more graphics processing units within the system.
However, FONG explicitly teaches wherein the collected system resource data further comprises data indicative of the availability of one or more graphics processing units within the system (Fig. 1. Paragraph [0026]-FONG discloses the AI platform (250) including a resource manager (258). Examples of resources include to a quantity of GPUs utilized, GPU types, etc. The profile manager (252) collects data associated with these resources and associated resource consumption from the corresponding ML model, and ML routines that are part of the ML model. In paragraph [0030]-FONG discloses the computing resource allocation may change across the iterations based on the requirements of the ML execution models as well as availability of the resources. In paragraph [0037]-FONG discloses possible execution directives include, setting limits of memory usage, and allocating resources for utilization. With respect to memory usage, limits may be set for application instances based on one or more previous iterations or historical data from one or more previous runs).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG in view of FONG and in further view of DWIVEDI of having a computer program product comprising: detecting an object in a two-dimensional image and identifying an object type of the detected object, with the teachings of FONG of having wherein the collected system resource data further comprises data indicative of the availability of one or more graphics processing units within the system.
Wherein WANG’s computer program product having wherein the collected system resource data further comprises data indicative of the availability of one or more graphics processing units within the system.
The motivation behind the modification would have been to obtain a computer program product that improve the functionality and operation of an artificial intelligence platform to model ML, resource usage, and the accuracy, quality and robustness of 3D models, since both WANG and FONG concern machine learning. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images using a learning model, while FONG provides systems and methods that improve the functionality and operation of an artificial intelligence platform to model ML application execution and resource usage, including efficient and effective ML application performance and resource management. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and FONG et al. (US 20210191759 A1), Abstract and paragraph [0077].
Claims 5, 12 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over WANG et al. (US 20110188780 A1), hereinafter referenced as WANG and in view of FONG et al. (US 20210191759 A1), hereinafter referenced as FONG and in further view of DWIVEDI et al. (US 20210366166 A1), hereinafter referenced as DWIVEDI and in further view of GALLO et al. (US 20230137403 A1), hereinafter referenced as GALLO.
Regarding claim 5, WANG in view of FONG and in further view of DWIVEDI explicitly teach the method of claim 1, WANG in view of FONG fails to explicitly teach wherein adapting the selected algorithm based on at least a portion of the collected system resource data further comprises adapting the one or more parameters of the selected algorithm to execute on respective portions of system resources associated with the one or more accelerators.
However, DWIVEDI explicitly teaches wherein adapting the selected algorithm (Fig. 2. Paragraph [0014]-DWIVEDI discloses the technique generates an optimal weighted execution sequence of available reconstruction algorithms using a multi-processor unit. The techniques can be parameterized to generate a weighted sequence that provides the best reconstructed image given time and computing resource. In paragraph [0015]-DWIVEDI discloses the technique dynamically determines the optimal weighted sequence for two or more algorithms by testing out multiple weighted sequences during reconstruction and choosing the best sequence. The technique tests multiple weighted sequences in parallel using the parallel processing capability of the multi-processor unit) based on at least a portion of the collected system resource data further comprises adapting the one or more parameters of the selected algorithm (Fig. 2. Paragraph [0023]-DWIVEDI discloses the processor 210s may select them from the available reconstruction algorithms based on factors such as: characteristics of available reconstruction algorithms, e.g., processing parameters that the algorithms enhance/reduce, iteration times of the algorithms, points of diminishing return for the algorithms (how effective each reconstruction algorithm is when it is iterated a certain number of times), and the preferred processing parameters (wherein the preferred processing parameters include at least one of a preferred performance parameter such as a preferred amount of time or computing resource for reconstruction, or a preferred visual parameter). Further in paragraph [0028]-DWIVEDI discloses the weighted execution sequences are determined based on various factors such as (without limitation) the preferred processing parameter, the characteristics of the selected reconstruction algorithms, and the time and computing constraints. The time and computing constraints represent the resources available to complete the request, computation time requirement, and available processing power) to execute on respective portions of system resources (Fig. 2. Paragraph [0029]-DWIVEDI discloses when the weighted execution sequences are determined, the selected algorithms are executed on the representative data according to the weighted execution sequences. The weighted execution sequences are executed in parallel using the multiple processors 210s of the multi-processor unit 200 for maximum efficiency and speed. Each of the weighted execution sequences may be executed on a different processor 210 of the multi-processor unit 200, or the weighted execution sequences may be executed one at a time using all the multiple processors 210s) associated with the one or more accelerators (Fig. 2, #110 and 210-N, called multi-processor unit and a processor, respectively. Paragraph [0017 and 0019]. Further in paragraph [0017]-DWIVEDI discloses the multi-processor unit 110 is configured to determine an optimal weighted execution sequence of reconstruction algorithms for processing image data from one or more of the devices. The multi-processor unit 110 may be any processing unit that is capable of parallel-processing, such as a graphics processing unit (GPU), a central processing unit (CPU), or a combination of both CPU and GPU)
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG in view of FONG and in further view of DWIVEDI of having a method comprising: detecting an object in a two-dimensional image and identifying an object type of the detected object, with the teachings of DWIVEDI of having wherein adapting the selected algorithm based on at least a portion of the collected system resource data further comprises adapting the one or more parameters of the selected algorithm to partition the selected algorithm to execute on respective portions of system resources associated with the one or more accelerators.
Wherein WANG’s method having wherein adapting the selected algorithm based on at least a portion of the collected system resource data further comprises adapting the one or more parameters of the selected algorithm to partition the selected algorithm to execute on respective portions of system resources associated with the one or more accelerators.
The motivation behind the modification would have been to obtain a method that improves accuracy, quality and robustness of 3D models and reconstruction, since both WANG and DWIVEDI concern systems and methods for image reconstruction. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images, while DWIVEDI provides systems and methods for improving the image reconstruction process. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and DWIVEDI et al. (US 20210366166 A1), Abstract and Paragraph [0057].
Although DWIVEDI explicitly teaches wherein adapting the selected algorithm based on at least a portion of the collected system resource data further comprises adapting the one or more parameters of the selected algorithm to execute on respective portions of system resources associated with the one or more accelerators.
WANG in view of FONG and in further view of DWIVEDI fail to explicitly teach partition the selected algorithm to execute on respective portions of system resources associated with the one or more accelerators.
However, GALLO explicitly teaches partition the selected algorithm to execute on respective portions of system resources associated with the one or more accelerators (Fig. 1. Paragraph [0051]-GALLO discloses volume renderer 312 (e.g., a neural radiance field (NeRF)-based volume renderer) is used to generate a sequence of images or video frames of an object or person in various poses. In paragraph [0121]-GALLO discloses a virtualized graphics execution environment is presented in which resources of graphics processing engines 1231-1232, N are shared with multiple applications or virtual machines (VMs). Resources may be subdivided into “slices” which are allocated to different VMs and/or applications based on processing requirements and priorities associated with VMs and/or applications. In paragraph [0164]-GALLO discloses graphics core 1500 includes a shared instruction cache 1502, a texture unit 1518, and a cache/shared memory 1520 that are common to execution resources within graphics core 1500. Graphics core 1500 can include multiple slices 1501A-1501N or partition for each core, and a graphics processor can include multiple instances of graphics core 1500)
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG in view of FONG and in further view of DWIVEDI of having a method comprising: detecting an object in a two-dimensional image and identifying an object type of the detected object, with the teachings of GALLO having partition the selected algorithm to execute on respective portions of system resources associated with the one or more accelerators.
Wherein WANG’s method having wherein adapting the selected algorithm based on at least a portion of the collected system resource data further comprises adapting the one or more parameters of the selected algorithm to partition the selected algorithm to execute on respective portions of system resources associated with the one or more accelerators.
The motivation behind the modification would have been to obtain a method that improves the speed, accuracy, quality and robustness of generating and rendering 3D models, since both WANG and GALLO concern systems and methods for object detection and 2D-3D image conversion. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images, while GALLO provides systems and methods for improving rendering speed with NERF based models and training speed of deep neural networks. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and GALLO et al. (US 20230137403 A1), Abstract and Paragraph [0057 and 0168].
Regarding claim 12, WANG in view of FONG and in further view of DWIVEDI explicitly teaches the apparatus of claim 8, WANG in view of FONG fails to explicitly teach wherein adapting the selected algorithm based on at least a portion of the collected system resource data further comprises adapting the one or more parameters of the selected algorithm to execute on respective portions of system resources associated with the one or more accelerators.
However, DWIVEDI explicitly teaches wherein adapting the selected algorithm (Fig. 2. Paragraph [0014]-DWIVEDI discloses the technique generates an optimal weighted execution sequence of available reconstruction algorithms using a multi-processor unit. The techniques can be parameterized to generate a weighted sequence that provides the best reconstructed image given time and computing resource. In paragraph [0015]-DWIVEDI discloses the technique dynamically determines the optimal weighted sequence for two or more algorithms by testing out multiple weighted sequences during reconstruction and choosing the best sequence. The technique tests multiple weighted sequences in parallel using the parallel processing capability of the multi-processor unit) based on at least a portion of the collected system resource data (Fig. 2. Paragraph [0023]-DWIVEDI discloses the processor 210s may select them from the available reconstruction algorithms based on factors such as: characteristics of available reconstruction algorithms, e.g., processing parameters that the algorithms enhance/reduce, iteration times of the algorithms, points of diminishing return for the algorithms (how effective each reconstruction algorithm is when it is iterated a certain number of times), and the preferred processing parameters (wherein the preferred processing parameters include at least one of a preferred performance parameter such as a preferred amount of time or computing resource for reconstruction, or a preferred visual parameter). Further in paragraph [0028]-DWIVEDI discloses the weighted execution sequences are determined based on various factors such as (without limitation) the preferred processing parameter, the characteristics of the selected reconstruction algorithms, and the time and computing constraints. The time and computing constraints represent the resources available to complete the request, computation time requirement, and available processing power) further comprises adapting the one or more parameters of the selected algorithm to execute on respective portions of system resources (Fig. 2. Paragraph [0029]-DWIVEDI discloses when the weighted execution sequences are determined, the selected algorithms are executed on the representative data according to the weighted execution sequences. The weighted execution sequences are executed in parallel using the multiple processors 210s of the multi-processor unit 200 for maximum efficiency and speed. Each of the weighted execution sequences may be executed on a different processor 210 of the multi-processor unit 200, or the weighted execution sequences may be executed one at a time using all the multiple processors 210s) associated with the one or more accelerators (Fig. 2, #110 and 210-N, called multi-processor unit and a processor, respectively. Paragraph [0017 and 0019]. Further in paragraph [0017]-DWIVEDI discloses the multi-processor unit 110 is configured to determine an optimal weighted execution sequence of reconstruction algorithms for processing image data from one or more of the devices. The multi-processor unit 110 may be any processing unit that is capable of parallel-processing, such as a graphics processing unit (GPU), a central processing unit (CPU), or a combination of both CPU and GPU).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG in view of FONG and in further view of DWIVEDI of having an apparatus comprising: detecting an object in a two-dimensional image and identifying an object type of the detected object, with the teachings of DWIVEDI of having wherein adapting the selected algorithm based on at least a portion of the collected system resource data further comprises adapting the one or more parameters of the selected algorithm to execute on respective portions of system resources associated with the one or more accelerators.
Wherein WANG’s apparatus having wherein adapting the selected algorithm based on at least a portion of the collected system resource data further comprises adapting the one or more parameters of the selected algorithm to execute on respective portions of system resources associated with the one or more accelerators.
The motivation behind the modification would have been to obtain an apparatus that improves accuracy, quality and robustness of 3D models and reconstruction, since both WANG and DWIVEDI concern systems and methods for image reconstruction. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images, while DWIVEDI provides systems and methods for improving the image reconstruction process. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and DWIVEDI et al. (US 20210366166 A1), Abstract and Paragraph [0057].
Although DWIVEDI explicitly teaches wherein adapting the selected algorithm based on at least a portion of the collected system resource data further comprises adapting the one or more parameters of the selected algorithm to execute on respective portions of system resources associated with the one or more accelerators.
WANG in view of FONG and in further view of DWIVEDI fail to explicitly teach partition the selected algorithm to execute on respective portions of system resources associated with the one or more accelerators.
However, GALLO explicitly teaches partition the selected algorithm to execute on respective portions of system resources associated with the one or more accelerators (Fig. 1. Paragraph [0051]-GALLO discloses volume renderer 312 (e.g., a neural radiance field (NeRF)-based volume renderer) is used to generate a sequence of images or video frames of an object or person in various poses. In paragraph [0121]-GALLO discloses a virtualized graphics execution environment is presented in which resources of graphics processing engines 1231-1232, N are shared with multiple applications or virtual machines (VMs). Resources may be subdivided into “slices” which are allocated to different VMs and/or applications based on processing requirements and priorities associated with VMs and/or applications. In paragraph [0164]-GALLO discloses graphics core 1500 includes a shared instruction cache 1502, a texture unit 1518, and a cache/shared memory 1520 that are common to execution resources within graphics core 1500. Graphics core 1500 can include multiple slices 1501A-1501N or partition for each core, and a graphics processor can include multiple instances of graphics core 1500).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG in view of FONG and in further view of DWIVEDI of having an apparatus comprising: detecting an object in a two-dimensional image and identifying an object type of the detected object, with the teachings of GALLO of having partition the selected algorithm to execute on respective portions of system resources associated with the one or more accelerators.
Wherein WANG’s apparatus having partition the selected algorithm to execute on respective portions of system resources associated with the one or more accelerators.
The motivation behind the modification would have been to obtain an apparatus that improves the speed, accuracy, quality and robustness of generating and rendering 3D models, since both WANG and GALLO concern systems and methods for object detection and 2D-3D image conversion. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images, while GALLO provides systems and methods for improving rendering speed with NERF based models and training speed of deep neural networks. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and GALLO et al. (US 20230137403 A1), Abstract and Paragraph [0057 and 0168].
Regarding claim 18, WANG in view of FONG and in further view of DWIVEDI explicitly teaches the computer program product of claim 15, WANG in view of FONG fails to explicitly teach wherein adapting the selected algorithm based on at least a portion of the collected system resource data further comprises adapting the one or more parameters of the selected algorithm to execute on respective portions of system resources associated with the one or more accelerators.
DWIVEDI explicitly teaches wherein adapting the selected algorithm (Fig. 2. Paragraph [0014]-DWIVEDI discloses the technique generates an optimal weighted execution sequence of available reconstruction algorithms using a multi-processor unit. The techniques can be parameterized to generate a weighted sequence that provides the best reconstructed image given time and computing resource. In paragraph [0015]-DWIVEDI discloses the technique dynamically determines the optimal weighted sequence for two or more algorithms by testing out multiple weighted sequences during reconstruction and choosing the best sequence. The technique tests multiple weighted sequences in parallel using the parallel processing capability of the multi-processor unit) based on at least a portion of the collected system resource data (Fig. 2. Paragraph [0023]-DWIVEDI discloses the processor 210s may select them from the available reconstruction algorithms based on factors such as: characteristics of available reconstruction algorithms, e.g., processing parameters that the algorithms enhance/reduce, iteration times of the algorithms, points of diminishing return for the algorithms (how effective each reconstruction algorithm is when it is iterated a certain number of times), and the preferred processing parameters (wherein the preferred processing parameters include at least one of a preferred performance parameter such as a preferred amount of time or computing resource for reconstruction, or a preferred visual parameter). Further in paragraph [0028]-DWIVEDI discloses the weighted execution sequences are determined based on various factors such as (without limitation) the preferred processing parameter, the characteristics of the selected reconstruction algorithms, and the time and computing constraints. The time and computing constraints represent the resources available to complete the request, computation time requirement, and available processing power) further comprises adapting the one or more parameters of the selected algorithm to execute on respective portions of system resources (Fig. 2. Paragraph [0029]-DWIVEDI discloses when the weighted execution sequences are determined, the selected algorithms are executed on the representative data according to the weighted execution sequences. The weighted execution sequences are executed in parallel using the multiple processors 210s of the multi-processor unit 200 for maximum efficiency and speed. Each of the weighted execution sequences may be executed on a different processor 210 of the multi-processor unit 200, or the weighted execution sequences may be executed one at a time using all the multiple processors 210s) associated with the one or more accelerators (Fig. 2, #110 and 210-N, called multi-processor unit and a processor, respectively. Paragraph [0017 and 0019]. Further in paragraph [0017]-DWIVEDI discloses the multi-processor unit 110 is configured to determine an optimal weighted execution sequence of reconstruction algorithms for processing image data from one or more of the devices. The multi-processor unit 110 may be any processing unit that is capable of parallel-processing, such as a graphics processing unit (GPU), a central processing unit (CPU), or a combination of both CPU and GPU).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG in view of FONG and in further view of DWIVEDI of having a computer program product stored on a non-transitory computer-readable medium and comprising machine executable instructions, the machine executable instructions, when executed, causing a processing device to perform steps, with the teachings of DWIVEDI of having wherein adapting the selected algorithm based on at least a portion of the collected system resource data further comprises adapting the one or more parameters of the selected algorithm to execute on respective portions of system resources associated with the one or more accelerators.
Wherein WANG’s computer program product having wherein adapting the selected algorithm based on at least a portion of the collected system resource data further comprises adapting the one or more parameters of the selected algorithm to execute on respective portions of system resources associated with the one or more accelerators.
The motivation behind the modification would have been to obtain a computer program product improves accuracy, quality and robustness of 3D models and reconstruction, since both WANG and DWIVEDI concern systems and methods for image reconstruction. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images, while DWIVEDI provides systems and methods for improving the image reconstruction process. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and DWIVEDI et al. (US 20210366166 A1), Abstract and Paragraph [0057].
Although DWIVEDI explicitly teaches wherein adapting the selected algorithm based on at least a portion of the collected system resource data further comprises adapting the one or more parameters of the selected algorithm to execute on respective portions of system resources associated with the one or more accelerators.
WANG in view of FONG and in further view of DWIVEDI fail to explicitly teach partition the selected algorithm to execute on respective portions of system resources associated with the one or more accelerators.
However, GALLO explicitly teaches partition the selected algorithm to execute on respective portions of system resources associated with the one or more accelerators (Fig. 1. Paragraph [0051]-GALLO discloses volume renderer 312 (e.g., a neural radiance field (NeRF)-based volume renderer) is used to generate a sequence of images or video frames of an object or person in various poses. In paragraph [0121]-GALLO discloses a virtualized graphics execution environment is presented in which resources of graphics processing engines 1231-1232, N are shared with multiple applications or virtual machines (VMs). Resources may be subdivided into “slices” which are allocated to different VMs and/or applications based on processing requirements and priorities associated with VMs and/or applications. In paragraph [0164]-GALLO discloses graphics core 1500 includes a shared instruction cache 1502, a texture unit 1518, and a cache/shared memory 1520 that are common to execution resources within graphics core 1500. Graphics core 1500 can include multiple slices 1501A-1501N or partition for each core, and a graphics processor can include multiple instances of graphics core 1500).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG in view of FONG and in further view of DWIVEDI of having a computer program product stored on a non-transitory computer-readable medium and comprising machine executable instructions, the machine executable instructions, when executed, causing a processing device to perform steps, with the teachings of GALLO of having partition the selected algorithm to execute on respective portions of system resources associated with the one or more accelerators.
Wherein WANG’s computer program product having wherein adapting the selected algorithm based on at least a portion of the collected system resource data further comprises adapting the one or more parameters of the selected algorithm to partition the selected algorithm to execute on respective portions of system resources associated with the one or more accelerators.
The motivation behind the modification would have been to obtain a computer program product that improves the speed, accuracy, quality and robustness of generating and rendering 3D models, since both WANG and GALLO concern systems and methods for object detection and 2D-3D image conversion. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images, while GALLO provides systems and methods for improving rendering speed with NERF based models and training speed of deep neural networks. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and GALLO et al. (US 20230137403 A1), Abstract and Paragraph [0057 and 0168].
Claims 7, 14, 20 and 27-29 are rejected under 35 U.S.C. 103 as being unpatentable over WANG et al. (US 20110188780 A1), hereinafter referenced as WANG and in view of FONG et al. (US 20210191759 A1), hereinafter referenced as FONG and in further view of DWIVEDI et al. (US 20210366166 A1), hereinafter referenced as DWIVEDI and in further view of WANG et al. (US 20210343054 A1), hereinafter referenced as WANG (2021).
Regarding claim 7, WANG in view of FONG and in further view of DWIVEDI explicitly teach the method of claim 1, although WANG explicitly teaches further comprising selecting a algorithm to transform the two-dimensional image into the three-dimensional model (Fig. 9. Paragraph [0029]-WANG discloses based on a result of the content analysis, e.g., the above-described image categorization/subcategorization and/or image object identification, a corresponding 2D-3D image conversion method may be chosen to generate a 3D image. Applying the chosen method to the 2D image, a 3D image can be generated (step 208)).
WANG in view of FONG fail to explicitly teach further comprising selecting a default algorithm to transform the two-dimensional image into the three-dimensional model when no algorithm is selected from the plurality of algorithms or when the testing of the adapted algorithm is unsuccessful.
However, WANG (2021) explicitly teaches further comprising selecting a default algorithm to transform the image into the model (Fig. 5. Paragraph [0065]-WANG discloses the algorithm determination module 404 may be configured to determine one or more reconstruction-related algorithms (wherein reconstruction-related algorithms are stored/managed in a container, which occupy a resource within a computing device and the algorithms may be selected automatically or by a user and include one or more of a plurality of sub-algorithms, pre-processing algorithms, image creating algorithms, and/or post-processing algorithms). Please also read paragraph [0066]) when no algorithm is selected from the plurality of algorithms (Fig. 7. Paragraph [0071]-WANG discloses the existing container or the newly generated container may contain any reconstruction-related algorithm, such as a default reconstruction-related algorithm, and the container determination module 406 may replace the default reconstruction-related algorithm with the reconstruction-related algorithm of the one or more reconstruction-related algorithms. In response to the reconstruction task, the algorithm determination module 404 may determine the reconstruction-related algorithm(s). For each of the reconstruction-related algorithms, the container determination module 406 may determine whether there is an existing container containing the reconstruction-related algorithm. In response to determining that there is no existing container containing the reconstruction-related algorithm, the algorithm determination module 404 may generate a new container for the reconstruction-related algorithm. Please also read paragraph [0067, 0091, 0096 and 0098]) or when the testing of the adapted algorithm is unsuccessful (Fig. 5. Paragraph [0091]-WANG discloses in response to determining that the target image does not satisfy the preset condition, the processing device 140 may generate a new target image automatically or in response to a user instruction. The user instruction may include replacing a portion of the one or more reconstruction-related algorithms by one or more new reconstruction-related algorithms. The one or more reconstruction-related algorithms may include one or more pre-processing algorithms, one or more image creating algorithms, one or more post-processing algorithms, etc. The processing device 140 may replace containers containing the one or more image creating algorithms by new containers containing new image creating algorithms. Alternatively, the processing device 140 may replace the one or more image creating algorithms with the new image creating algorithms in the containers to update the containers. Further, the processing device 140 may update the reconstruction flow based on the new image creating algorithms and their corresponding containers).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG in view of FONG and in further view of DWIVEDI of having a method comprising: detecting an object in a two-dimensional image and identifying an object type of the detected object, with the teachings of WANG (2021) of having selecting a default algorithm to transform the image into the model when no algorithm is selected from the plurality of algorithms or when the testing of the adapted algorithm is unsuccessful.
Wherein WANG’s method having further comprising selecting a default algorithm to transform the two-dimensional image into the three-dimensional model when no algorithm is selected from the plurality of algorithms or when the testing of the adapted algorithm is unsuccessful.
The motivation behind the modification would have been to obtain a method that improves the accuracy, quality and computational efficiency of image/model generation reconstructions, since both WANG and WANG (2021) concern systems and methods for image processing and generation. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images, while WANG (2021)’s systems and methods allow for the reconstruction flow to be dynamically divided and adjusted by adding/replacing/deleting algorithm(s) contained one or more containers, which improves the reconstruction efficiency. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and WANG et al. (US 20210343054 A1), Abstract and Paragraph [0091].
Regarding claim 14, WANG in view of FONG and in further view of DWIVEDI explicitly teach the apparatus of claim 8, although WANG explicitly teaches wherein the apparatus is further configured to select a default algorithm to transform the two-dimensional image into the three- dimensional model (Fig. 9. Paragraph [0029]-WANG discloses based on a result of the content analysis, e.g., the above-described image categorization/subcategorization and/or image object identification, a corresponding 2D-3D image conversion method may be chosen to generate a 3D image. Applying the chosen method to the 2D image, a 3D image can be generated (step 208)).
WANG in view of FONG fail to explicitly teach wherein the apparatus is further configured to select a default algorithm to transform the two-dimensional image into the three-dimensional model when no algorithm is selected from the plurality of algorithms or when the testing of the adapted algorithm is unsuccessful.
However, WANG (2021) explicitly teaches wherein the apparatus is further configured to select a default algorithm to transform the image into the model (Fig. 5. Paragraph [0065]-WANG discloses the algorithm determination module 404 may be configured to determine one or more reconstruction-related algorithms (wherein reconstruction-related algorithms are stored/managed in a container, which occupy a resource within a computing device and the algorithms may be selected automatically or by a user and include one or more of a plurality of sub-algorithms, pre-processing algorithms, image creating algorithms, and/or post-processing algorithms). Please also read paragraph [0066]) when no algorithm is selected from the plurality of algorithms (Fig. 5. Paragraph [0071]-WANG discloses the existing container or the newly generated container may contain any reconstruction-related algorithm, such as a default reconstruction-related algorithm, and the container determination module 406 may replace the default reconstruction-related algorithm with the reconstruction-related algorithm of the one or more reconstruction-related algorithms. In response to the reconstruction task, the algorithm determination module 404 may determine the reconstruction-related algorithm(s). For each of the reconstruction-related algorithms, the container determination module 406 may determine whether there is an existing container containing the reconstruction-related algorithm. In response to determining that there is no existing container containing the reconstruction-related algorithm, the algorithm determination module 404 may generate a new container for the reconstruction-related algorithm. Please also read paragraph [0067, 0091, 0096 and 0098]) or when the testing of the adapted algorithm is unsuccessful (Fig. 5. Paragraph [0091]-WANG discloses in response to determining that the target image does not satisfy the preset condition, the processing device 140 may generate a new target image automatically or in response to a user instruction. The user instruction may include replacing a portion of the one or more reconstruction-related algorithms by one or more new reconstruction-related algorithms. The one or more reconstruction-related algorithms may include one or more pre-processing algorithms, one or more image creating algorithms, one or more post-processing algorithms, etc. The processing device 140 may replace containers containing the one or more image creating algorithms by new containers containing new image creating algorithms. Alternatively, the processing device 140 may replace the one or more image creating algorithms with the new image creating algorithms in the containers to update the containers. Further, the processing device 140 may update the reconstruction flow based on the new image creating algorithms and their corresponding containers).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG in view of FONG and in further view of DWIVEDI of having an apparatus comprising: detecting an object in a two-dimensional image and identifying an object type of the detected object, with the teachings of WANG (2021) of having wherein the apparatus is further configured to select a default algorithm to transform the image into the model when no algorithm is selected from the plurality of algorithms or when the testing of the adapted algorithm is unsuccessful.
Wherein WANG’s apparatus having wherein the apparatus is further configured to select a default algorithm to transform the two-dimensional image into the three-dimensional model when no algorithm is selected from the plurality of algorithms or when the testing of the adapted algorithm is unsuccessful.
The motivation behind the modification would have been to obtain an apparatus that improves the accuracy, quality and computational efficiency of image/model generation reconstructions, since both WANG and WANG (2021) concern systems and methods for image processing and generation. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images, while WANG (2021)’s systems and methods allow for the reconstruction flow to be dynamically divided and adjusted by adding/replacing/deleting algorithm(s) contained one or more containers, which improves the reconstruction efficiency. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and WANG et al. (US 20210343054 A1), Abstract and Paragraph [0091].
Regarding claim 20, WANG in view of FONG and in further view of DWIVEDI explicitly teaches the computer program product of claim 17, although WANG explicitly teaches further comprising selecting a algorithm to transform the two-dimensional image into the three-dimensional model (Fig. 9. Paragraph [0029]-WANG discloses based on a result of the content analysis, e.g., the above-described image categorization/subcategorization and/or image object identification, a corresponding 2D-3D image conversion method may be chosen to generate a 3D image. Applying the chosen method to the 2D image, a 3D image can be generated (step 208)).
WANG in view of FONG fail to explicitly teach selecting a default algorithm to transform the two-dimensional image into the three-dimensional model when no algorithm is selected from the plurality of algorithms or when the testing of the adapted algorithm is unsuccessful.
However, WANG (2021) explicitly teaches selecting a default algorithm to transform the image into the model (Fig. 5. Paragraph [0065]-WANG discloses the algorithm determination module 404 may be configured to determine one or more reconstruction-related algorithms (wherein reconstruction-related algorithms are stored/managed in a container, which occupy a resource within a computing device and the algorithms may be selected automatically or by a user and include one or more of a plurality of sub-algorithms, pre-processing algorithms, image creating algorithms, and/or post-processing algorithms). Please also read paragraph [0066]) when no algorithm is selected from the plurality of algorithms (Fig. 5. Paragraph [0071]-WANG discloses the existing container or the newly generated container may contain any reconstruction-related algorithm, such as a default reconstruction-related algorithm, and the container determination module 406 may replace the default reconstruction-related algorithm with the reconstruction-related algorithm of the one or more reconstruction-related algorithms. In response to the reconstruction task, the algorithm determination module 404 may determine the reconstruction-related algorithm(s). For each of the reconstruction-related algorithms, the container determination module 406 may determine whether there is an existing container containing the reconstruction-related algorithm. In response to determining that there is no existing container containing the reconstruction-related algorithm, the algorithm determination module 404 may generate a new container for the reconstruction-related algorithm. Please also read paragraph [0067, 0091, 0096 and 0098]) or when the testing of the adapted algorithm is unsuccessful (Fig. 5. Paragraph [0091]-WANG discloses in response to determining that the target image does not satisfy the preset condition, the processing device 140 may generate a new target image automatically or in response to a user instruction. The user instruction may include replacing a portion of the one or more reconstruction-related algorithms by one or more new reconstruction-related algorithms. The one or more reconstruction-related algorithms may include one or more pre-processing algorithms, one or more image creating algorithms, one or more post-processing algorithms, etc. The processing device 140 may replace containers containing the one or more image creating algorithms by new containers containing new image creating algorithms. Alternatively, the processing device 140 may replace the one or more image creating algorithms with the new image creating algorithms in the containers to update the containers. Further, the processing device 140 may update the reconstruction flow based on the new image creating algorithms and their corresponding containers).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG in view of FONG and in further view of DWIVEDI of having a computer program product stored on a non-transitory computer-readable medium and comprising machine executable instructions, the machine executable instructions, when executed, causing a processing device to perform steps, with the teachings of WANG (2021) of having further comprising selecting a default algorithm to transform the image into the model when no algorithm is selected from the plurality of algorithms or when the testing of the adapted algorithm is unsuccessful.
Wherein WANG’s computer program product further comprising selecting a default algorithm to transform the two-dimensional image into the three-dimensional model when no algorithm is selected from the plurality of algorithms or when the testing of the adapted algorithm is unsuccessful.
The motivation behind the modification would have been to obtain a computer program product that improves the accuracy, quality and computational efficiency of image/model generation reconstructions, since both WANG and WANG (2021) concern systems and methods for image processing and generation. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images, while WANG (2021)’s systems and methods allow for the reconstruction flow to be dynamically divided and adjusted by adding/replacing/deleting algorithm(s) contained one or more containers, which improves the reconstruction efficiency. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and WANG et al. (US 20210343054 A1), Abstract and Paragraph [0091].
Regarding claim 27, WANG in view of FONG and in further view of DWIVEDI explicitly teaches the method of claim 1, WANG in view of FONG fails to explicitly teach wherein the selecting the algorithm from the plurality of algorithms is based on an algorithm map configured to maintain a pre-defined correspondence between a plurality of object types and the plurality of algorithms.
However, WANG (2021) explicitly teaches wherein the selecting the algorithm from the plurality of algorithms is based on an algorithm map configured to maintain a pre-defined correspondence between a plurality of object types and the plurality of algorithms (Fig. 5. Paragraph [0065]-WANG discloses the algorithm determination module 404 may determine one or more reconstruction-related algorithms based on the raw data. The one or more reconstruction-related algorithms may be determined based on the object under examination. Merely by way of example, the algorithm determining module 404 may determine the Bone Correction in the case that the object under the examination is the head. Please also read paragraph [0065]).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG in view of FONG and in further view of DWIVEDI of having a method comprising: detecting an object in a two-dimensional image and identifying an object type of the detected object, with the teachings of WANG (2021) of having system resource data comprising data indicative of one or more accelerators within a system including one or more accelerator types, utilization data and availability data associated with the one or more accelerators of the system;
Wherein WANG’s method having system resource data comprising data indicative of one or more accelerators within a system including one or more accelerator types, utilization data and availability data associated with the one or more accelerators of the system;
The motivation behind the modification would have been to obtain a method that improves the accuracy, quality and computational efficiency of image/model generation reconstructions, since both WANG and WANG (2021) concern systems and methods for image processing and generation. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images, while WANG (2021)’s systems and methods allow for the reconstruction flow to be dynamically divided and adjusted by adding/replacing/deleting algorithm(s) contained one or more containers, which improves the reconstruction efficiency. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and WANG et al. (US 20210343054 A1), Abstract and Paragraph [0091].
Regarding claim 28, WANG in view of FONG and in further view of DWIVEDI explicitly teach the apparatus of claim 8, WANG in view of FONG fails to explicitly teach wherein the selecting the algorithm from the plurality of algorithms is based on an algorithm map configured to maintain a pre-defined correspondence between a plurality of object types and the plurality of algorithms.
However, WANG (2021) explicitly teaches wherein the selecting the algorithm from the plurality of algorithms is based on an algorithm map configured to maintain a pre-defined correspondence between a plurality of object types and the plurality of algorithms (Fig. 5. Paragraph [0065]-WANG discloses the algorithm determination module 404 may determine one or more reconstruction-related algorithms based on the raw data. The one or more reconstruction-related algorithms may be determined based on the object under examination. Merely by way of example, the algorithm determining module 404 may determine the Bone Correction in the case that the object under the examination is the head. Please also read paragraph [0065]).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG in view of FONG and in further view of DWIVEDI of having an apparatus comprising: detecting an object in a two-dimensional image and identifying an object type of the detected object, with the teachings of WANG (2021) of having wherein the selecting the algorithm from the plurality of algorithms is based on an algorithm map configured to maintain a pre-defined correspondence between a plurality of object types and the plurality of algorithms.
Wherein WANG’s apparatus having wherein the selecting the algorithm from the plurality of algorithms is based on an algorithm map configured to maintain a pre-defined correspondence between a plurality of object types and the plurality of algorithms.
The motivation behind the modification would have been to obtain an apparatus that improves the accuracy, quality and computational efficiency of image/model generation reconstructions, since both WANG and WANG (2021) concern systems and methods for image processing and generation. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images, while WANG (2021)’s systems and methods allow for the reconstruction flow to be dynamically divided and adjusted by adding/replacing/deleting algorithm(s) contained one or more containers, which improves the reconstruction efficiency. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and WANG et al. (US 20210343054 A1), Abstract and Paragraph [0091].
Regarding claim 29, WANG in view of FONG and in further view of DWIVEDI explicitly teach the computer program product of claim 15, WANG in view of FONG fails to explicitly teach wherein the selecting the algorithm from the plurality of algorithms is based on an algorithm map configured to maintain a pre-defined correspondence between a plurality of object types and the plurality of algorithms.
However, WANG (2021) explicitly teaches wherein the selecting the algorithm from the plurality of algorithms is based on an algorithm map configured to maintain a pre-defined correspondence between a plurality of object types and the plurality of algorithms (Fig. 5. Paragraph [0065]-WANG discloses the algorithm determination module 404 may determine one or more reconstruction-related algorithms based on the raw data. The one or more reconstruction-related algorithms may be determined based on the object under examination. Merely by way of example, the algorithm determining module 404 may determine the Bone Correction in the case that the object under the examination is the head. Please also read paragraph [0065]).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG in view of FONG and in further view of DWIVEDI of having a computer program product comprising: detecting an object in a two-dimensional image and identifying an object type of the detected object, with the teachings of WANG (2021) of having wherein the selecting the algorithm from the plurality of algorithms is based on an algorithm map configured to maintain a pre-defined correspondence between a plurality of object types and the plurality of algorithms.
Wherein WANG’s computer program product having wherein the selecting the algorithm from the plurality of algorithms is based on an algorithm map configured to maintain a pre-defined correspondence between a plurality of object types and the plurality of algorithms.
The motivation behind the modification would have been to obtain a computer program product that improves the accuracy, quality and computational efficiency of image/model generation reconstructions, since both WANG and WANG (2021) concern systems and methods for image processing and generation. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images, while WANG (2021)’s systems and methods allow for the reconstruction flow to be dynamically divided and adjusted by adding/replacing/deleting algorithm(s) contained one or more containers, which improves the reconstruction efficiency. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and WANG et al. (US 20210343054 A1), Abstract and Paragraph [0091].
Claims 22, 24, and 26 are rejected under 35 U.S.C. 103 as being unpatentable over WANG et al. (US 20110188780 A1), hereinafter referenced as WANG and in view of FONG et al. (US 20210191759 A1), hereinafter referenced as FONG and in further view of DWIVEDI et al. (US 20210366166 A1), hereinafter referenced as DWIVEDI and in further view of BROWN et al. (US 20230130281 A1), hereinafter referenced as BROWN.
Regarding claim 22, WANG in view of FONG and in further view of DWIVEDI explicitly teaches the method of claim 1, WANG in view of FONG fails to explicitly teach wherein the plurality of algorithms comprise a plurality of neural radiance fields algorithms.
However, BROWN explicitly teaches wherein the plurality of algorithms comprises a plurality of neural radiance fields algorithms (Fig. 1. Paragraph [0097]-BROWN discloses the foreground data can include a foreground output (e.g., neural radiance field data, which can include or be based on a five-dimensional function), which can include density data and color data. The object category model 200 can include a background model 206 (e.g., a background neural radiance field model) that is operable to output background data (e.g., neural radiance field data). In paragraph [0098]-BROWN discloses the systems and methods can include a FiG-NeRF architecture including foreground and background models as depicted in FIG. 2. The foreground model can include the deformation field 202 and template NeRF 204, and the background model 206 can include a template NeRF. In paragraph [0141]-BROWN discloses the systems and methods can include Neural Radiance Fields (NeRF). Moreover, the systems and methods can include Figure-Ground Neural Radiance Fields (FiG-NeRF), which can use two NeRF models to model the objects and background, respectively).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG in view of FONG and in further view of DWIVEDI of having a method comprising: detecting an object in a two-dimensional image and identifying an object type of the detected object, comprising: detecting an object in a two-dimensional image and identifying an object type of the detected object; and selecting, based on the identified object type of the detected object, an algorithm from a plurality of algorithms configured to transform a two-dimensional image into a three-dimensional model, with the teachings of BROWN of having wherein the plurality of algorithms comprise a plurality of neural radiance fields algorithms.
Wherein WANG’s method having wherein the plurality of algorithms comprise a plurality of neural radiance fields algorithms.
The motivation behind the modification would have been to obtain a method that improves the quality, efficiency and robustness of 3D models, since both WANG and BROWN concern systems and methods for 2D-3D image conversion. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images, while BROWN provides systems and methods that improve computational efficiency and improvements in the functioning of a computing system. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and BROWN et al. (US 20230130281 A1), Abstract and Paragraph [0061].
Regarding claim 24, WANG in view of FONG and in further view of DWIVEDI explicitly teaches the apparatus of claim 8, WANG in view of FONG fails to explicitly teach wherein the plurality of algorithms comprises a plurality of neural radiance fields algorithms.
However, BROWN explicitly teaches wherein the plurality of algorithms comprises a plurality of neural radiance fields algorithms (Fig. 1. Paragraph [0097]-BROWN discloses the foreground data can include a foreground output (e.g., neural radiance field data, which can include or be based on a five-dimensional function), which can include density data and color data. The object category model 200 can include a background model 206 (e.g., a background neural radiance field model) that is operable to output background data (e.g., neural radiance field data). In paragraph [0098]-BROWN discloses the systems and methods can include a FiG-NeRF architecture including foreground and background models as depicted in FIG. 2. The foreground model can include the deformation field 202 and template NeRF 204, and the background model 206 can include a template NeRF. In paragraph [0141]-BROWN discloses the systems and methods can include Neural Radiance Fields (NeRF). Moreover, the systems and methods can include Figure-Ground Neural Radiance Fields (FiG-NeRF), which can use two NeRF models to model the objects and background, respectively).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG in view of FONG and in further view of DWIVEDI of having an apparatus comprising: detecting an object in a two-dimensional image and identifying an object type of the detected object, comprising: at least one processor and at least one memory storing computer program instructions wherein, when the at least one processor executes the computer program instructions, the apparatus is configured to: detect an object in a two-dimensional image and identifying an object type of the detected object, with the teachings of BROWN of having wherein the plurality of algorithms comprise a plurality of neural radiance fields algorithms.
Wherein WANG’s apparatus having wherein the plurality of algorithms comprise a plurality of neural radiance fields algorithms.
The motivation behind the modification would have been to obtain an apparatus that improves the quality, efficiency and robustness of 3D models, since both WANG and BROWN concern systems and methods for 2D-3D image conversion. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images, while BROWN provides systems and methods that improve computational efficiency and improvements in the functioning of a computing system. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and BROWN et al. (US 20230130281 A1), Abstract and Paragraph [0061].
Regarding claim 26, WANG in view of FONG and in further view of DWIVEDI explicitly teaches the computer program product of claim 15, WANG in view of FONG fails to explicitly teach wherein the plurality of algorithms comprise a plurality of neural radiance fields algorithms.
However, BROWN explicitly teaches wherein the plurality of algorithms comprises a plurality of neural radiance fields algorithms (Fig. 1. Paragraph [0097]-BROWN discloses the foreground data can include a foreground output (e.g., neural radiance field data, which can include or be based on a five-dimensional function), which can include density data and color data. The object category model 200 can include a background model 206 (e.g., a background neural radiance field model) that is operable to output background data (e.g., neural radiance field data). In paragraph [0098]-BROWN discloses the systems and methods can include a FiG-NeRF architecture including foreground and background models as depicted in FIG. 2. The foreground model can include the deformation field 202 and template NeRF 204, and the background model 206 can include a template NeRF. In paragraph [0141]-BROWN discloses the systems and methods can include Neural Radiance Fields (NeRF). Moreover, the systems and methods can include Figure-Ground Neural Radiance Fields (FiG-NeRF), which can use two NeRF models to model the objects and background, respectively).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of WANG in view of FONG and in further view of DWIVEDI of having a computer program product stored on a non-transitory computer-readable medium and comprising machine executable instructions, the machine executable instructions, when executed, causing a processing device to perform steps, with the teachings of BROWN of having wherein the plurality of algorithms comprise a plurality of neural radiance fields algorithms..
Wherein WANG’s computer program product having wherein the plurality of algorithms comprise a plurality of neural radiance fields algorithms.
The motivation behind the modification would have been to obtain a computer program product that improves the quality, efficiency and robustness of 3D models, since both WANG and BROWN concern systems and methods for 2D-3D image conversion. Wherein WANG’s systems and methods provide an efficient way to obtain accurate 3D models from 2D images, while BROWN provides systems and methods that improve computational efficiency and improvements in the functioning of a computing system. Please see WANG et al. (US 20110188780 A1), Abstract and Paragraph [0003-0006] and BROWN et al. (US 20230130281 A1), Abstract and Paragraph [0061].
Conclusion
Listed below are the prior arts made of record and not relied upon but are considered pertinent to applicant`s disclosure.
MATTHEWS et al. (US 20240371081 A1)- Systems and methods for learning spaces of three-dimensional shape and appearance from datasets of single-view images can be utilized for generating view renderings of a variety of different objects and/or scenes. The systems and methods can be able to learn effectively from unstructured. “in-the-wild” data, without incurring the high cost of a full-image discriminator, and while avoiding problems such as mode-dropping that are inherent to adversarial methods.......................Please see Fig. 1-3. Abstract.
LIN et al. (US 20230230275 A1)- Provided are systems and methods that invert a trained NeRF model, which stores the structure of a scene or object, to estimate the 6D pose from an image taken with a novel view. 6D pose estimation has a wide range of applications, including visual localization and object pose estimation for robot manipulation.......................Please see Fig. 1-2. Abstract.
LV et al. (US 20220239844 A1)- In one embodiment, a method includes initializing latent codes respectively associated with times associated with frames in a training video of a scene captured by a camera. For each of the frames, a system (1) generates rendered pixel values for a set of pixels in the frame by querying NeRF using the latent code associated with the frame, a camera viewpoint associated with the frame, and ray directions associated with the set of pixels, and (2) updates the latent code associated with the frame and the NeRF based on comparisons between the rendered pixel values and original pixel values for the set of pixels. Once trained, the system renders output frames for an output video of the scene, wherein each output frame is rendered by querying the updated NeRF using one of the updated latent codes corresponding to a desired time associated with the output frame......................Please see Fig. 7. Abstract.
Rematas et al. (US 20230281913 A1)- Systems and methods for view synthesis and three-dimensional reconstruction can learn an environment by utilizing a plurality of images of an environment and depth data. The use of depth data can be helpful when the quantity of images and different angles may be limited. For example, large outdoor environments can be difficult to learn due to the size, the varying image exposures, and the limited variance in view direction changes. The systems and methods can leverage a plurality of panoramic images and corresponding lidar data to accurately learn a large outdoor environment to then generate view synthesis outputs and three-dimensional reconstruction outputs. Training may include the use of an exposure correction network to address lighting exposure differences between training images........................Please see Fig. 1-2. Abstract.
CHAI et al. (US 20240193855 A1)- A 3D-aware generative model for high-quality and controllable scene synthesis uses an abstract object-level representation (i.e., 3D bounding boxes without semantic annotation) as the scene layout prior, which is simple to obtain, general to describe various scene contents, and yet informative to disentangle objects and background. An overall layout for the scene is identified and then each object is located in the layout to facilitate the scene composition process. The object-level representation serves as an intuitive user control for scene editing. Based on such a prior, the system spatially disentangles the whole scene into object-centric generative radiance fields by learning on only 2D images with global-local discrimination. Once the model is trained, users can generate and edit a scene by explicitly controlling the camera and the layout of objects' bounding boxes.........................Please see Fig. 1-3. Abstract.
Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any extension fee pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner
should be directed to Aaron Bonansinga whose telephone number is (703) 756-5380 The examiner can normally be reached on Monday-Friday, 9:00 a.m. - 6:00 p.m. ET.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s
supervisor, Chineyere Wills-Burns can be reached by phone at (571) 272-9752. The fax phone number for the organization where this application or proceeding is assigned is (571) 273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/AARON TIMOTHY BONANSINGA/Examiner, Art Unit 2673
/CHINEYERE WILLS-BURNS/Supervisory Patent Examiner, Art Unit 2673
Prosecution Insights