DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
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 (i.e., changing from AIA to pre-AIA ) 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.
Joint Inventors
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 10/14/2025 is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Claim Rejections - 35 USC § 103
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, 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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1-6, 8, and 10 are rejected under 35 U.S.C. 103 as being unpatentable over Yu et al. (US20220261593, referred to as Yu) in view of Ratliff et al. (US12240112, referred to as Ratliff).
Regarding claim 1: Yu discloses: One or more processors, comprising: one or more circuits to: update, during a first update stage, a teacher model to generate first actions for a [geometric fabric] associated with a simulated autonomous machine of a simulation using state information of the simulation; ([0091] parameters of a teacher network 204 are updated by one or more systems using an exponential moving average, which is defined by a following equation, although any variation thereof can be utilized: 0,.-m0,+(l-m)0, where 0,, 0s, m denote parameters of a teacher network 204, a student network 206, and momentum, respectively. In at least one embodiment, updating parameters through an exponential moving average improves a stability of a teacher network 204 and enforces an output more consistent between iterations.) update, during a second update stage, a student model ([0089] teacher network 204 shares a same architecture as a student network 206, with an
addition of a conditional random fields (CRF) module.) to generate second actions for the [geometric fabric] using at least one rendered image of the simulation, the teacher model, and ([0067] one or more systems of a framework 106 for object detection 108, instance segmentation 110, and semantic correspondence 112 perform semantic correspondence 112 based on features and instance proposals output from a teacher network and/or a student network, and box proposals output from said student network to determine correspondence between objects of images 102, 104. In at least one embodiment, one or more systems of a framework 106 calculate loss by comparing outputs of a student network with outputs of a teacher network.) noised state information of the simulation; and ([0543] a visualization service may be used that may add image rendering effects, such as ray-tracing, rasterization, denoising, sharpening, etc., to add realism to two-dimensional (2D) and/or three-dimensional (3D) models.) control, using the student model [and the geometric fabric], a physical autonomous machine with respect to a physical object based at least on an image of an environment [including the physical autonomous machine and the physical object.] ([0061] a processor comprises one or more circuits to use one or more neural networks to segment one or more objects of one or more images 102, 104 based, at least in part, on one or more bounding boxes identified by said one or more neural networks. In at least one embodiment, said processor performs such operations as described in FIG. 2, which shows one or more training processes for a framework 106 for object detection 108, instance segmentation 110, and semantic correspondence 112. [0062] In at least one embodiment, a first image 102 and/or a second image 104 are images captured from one or more images and/or video capturing devices, such as one or more systems of a vehicle (e.g., autonomous vehicle, semi-autonomous vehicle) or other imaging system (e.g., robotic device, grasping device, medical imaging device, satellite imaging system).)
Yu does not explicitly disclose: geometric fabric … including the physical autonomous machine and the physical object.
Yu does not disclose the following limitations, however Ratliff, in an analogous field of endeavor teaches: geometric fabric … ([col. 5, lines 3-5] a computer system can be provided that generates a policy or geometric fabric that includes policy layers.) including the physical autonomous machine and the physical object. ([col. 49, lines 55-65] a unit may be used to generate a 3D map of an environment of vehicle 1200, including a distance estimate for all points in an image. In at least one embodiment, one or more of stereo camera(s) 1268 may include, without limitation, compact stereo vision sensor(s) that may include, without limitation, two camera lenses (one each on left and right) and an image processing chip that may measure distance from vehicle 1200 to target object and use generated information (e.g., metadata) to activate autonomous emergency braking and lane departure warning functions.)
Yu and Ratliff are analogous art to the claimed invention since they are from the similar field of neural network processing for pick and place robotic control. It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention, with a reasonable expectation for success, to modify the neural processing of Yu to enable the specific requirements for the imaging and network as described in Ratliff.
The motivation for modification would have been to provide the neural network processing disclosed in Yu with the particular processor taught in Ratliff for the purpose of improving training with a specific class of controller. Further, the motivation for including the machine and object in the processing data would be to make sure to only pull data when the most important parts of the control network are being imaged.
Regarding claim 2: The combination of Yu and Ratliff teaches: The one or more processors of claim 1,
Yu further discloses: wherein the student model comprises one or more transformer layers. ([0415] In at least one embodiment, neuron outputs 2606 of neurons 2602 in a first layer 2610 may be connected to neuron inputs 2604 of neurons 2602 in a second layer 2612. In at least one embodiment, layer 2610 may be referred to as a “feed-forward layer.” In at least one embodiment, each instance of neuron 2602 in an instance of first layer 2610 may fan out to each instance of neuron 2602 in second layer 2612.)
Regarding claim 3: The combination of Yu and Ratliff teaches: The one or more processors of claim 2,
Yu further discloses: wherein the student model comprises at least one recurrent neural network (RNN) layer and at least one fully-connected layer. ([0183] accelerator(s) 1114 could be used for targeted workloads (e.g., perception, convolutional neural networks (“CNNs”), recurrent neural networks (“RNNs”), etc.) that are stable enough to be amenable to acceleration. In at least one embodiment, a CNN may include a region-based or regional convolutional neural networks (“RCNNs”) and Fast RCNNs (e.g., as used for object detection) or other type of CNN.)
Regarding claim 4: The combination of Yu and Ratliff teaches: The one or more processors of claim 1,
Yu further discloses: wherein the one or more circuits are to: generate, during the second update phase, ([0089] teacher network 204 shares a same architecture as a student network 206, with an
addition of a conditional random fields (CRF) module.) an auxiliary loss based at least on a predicted position of a simulated object generated by the student model and a ground-truth object position derived from the simulation. ([0067] one or more systems of a framework 106 for object detection 108, instance segmentation 110, and semantic correspondence 112 perform semantic correspondence 112 based on features and instance proposals output from a teacher network and/or a student network, and box proposals output from said student network to determine correspondence between objects of images 102, 104. In at least one embodiment, one or more systems of a framework 106 calculate loss by comparing outputs of a student network with outputs of a teacher network.)
Regarding claim 5: The combination of Yu and Ratliff teaches: The one or more processors of claim 3,
Yu further discloses: wherein the one or more circuits are to: generate a loss for updating the student model ([0067] one or more systems of a framework 106 calculate loss by comparing outputs of a student network with outputs of a teacher network. In at least one embodiment, one or more systems of a framework 106 calculate loss by comparing outputs of a student network with training data 102, 104. In at least one embodiment, one or more systems of framework 106 update one or more parameters of a student network such that loss is minimized.) based at least on the auxiliary loss and a second loss generated using an output of the teacher model. ([0110] a system performing at least a part of process 700 includes executable code to determine 706 detection loss and auxiliary loss based at least in part on a student network. In at least one embodiment, a system determines detection loss by determining differences between bounding box proposals generated by a student network and bounding box annotations of training data. In at least one embodiment, a system determines auxiliary loss, also referred to as auxiliary softmax cross-entropy loss, by analyzing segmentation maps generated by a student network and bounding box annotations of training data.)
Regarding claim 6: The combination of Yu and Ratliff teaches: The one or more processors of claim 1,
Yu further discloses: wherein the one or more circuits are to: update the student model further based at least on [proprioception] data derived from the simulation. ([0067] one or more systems of a framework 106 calculate loss by comparing outputs of a student network with training data 102, 104. In at least one embodiment, one or more systems of framework 106 update one or more parameters of a student network such that loss is minimized.)
Yu does not explicitly disclose: proprioception data
Yu does not disclose the following limitations, however Ratliff, from an analogous field of endeavor, further teaches: proprioception data ([col. 39, lines 39-45] one or more motion parameters of the second
policy layer can be energized independent of the one or more motion parameters of the first policy layer. The one or more motion parameters can include trajectory motion parameters, acceleration motion parameters, velocity motion parameters, joint limitation parameters, redundancy parameters, directional and/or coordinate parameters, and so forth. [col. 5, lines 27-32] The user can use the UI to generate a second policy layer included in the policy. In at least one embodiment, the
second policy layer is to cause the machine to generate a second motion. The second motion can build on the first motion that the machine executes based on the first policy layer.)
As previously stated, Yu and Ratliff are analogous art to the claimed invention since they are from the similar field of neural network processing for pick and place robotic control. It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention, with a reasonable expectation for success, to modify the neural processing of Yu to enable the specific collection of proprioception data.
The motivation for modification would have been to provide the neural network processing disclosed in Yu with the particular processor taught in Ratliff for the purpose of providing controlled limits to joint mechanics.
Regarding claim 8: The combination of Yu and Ratliff teaches: The one or more processors of claim 1,
Yu further discloses: wherein the one or more circuits are to: update the teacher model according to an automatic domain randomization function. ([0090] a teacher network 204 and a student network 206 form a self-distillation framework to further improve segmentation quality. In at least one embodiment, a student network 206 is a learnable network that performs instance segmentation. In at least one embodiment, a teacher network 204 is a temporally-consistent network guided by one or more CRF modules to generate pseudo-labels from images to generate less noisy and sharper instance segmentation)
Regarding claim 10: The combination of Yu and Ratliff teaches: The one or more processors of claim 1,
Yu further discloses: wherein the one or more circuits are to: update, during the second update stage, the student model to generate second actions for the [geometric fabric] using a plurality of rendered images of the simulation. ([0251] training data may be generated by vehicles, and/or may be generated in a simulation (e.g., using a game engine). In at least one embodiment, any amount of training data is tagged (e.g., where associated neural network benefits from supervised learning) and/or undergoes other pre-processing. In at least one embodiment, any amount of training data is not tagged and/or pre-processed (e.g., where associated neural network does not require supervised learning). In at least one embodiment, once machine learning models are trained, machine learning models may be used by vehicles (e.g., transmitted to vehicles over network(s) 1190), and/or machine learning models may be used by server(s) 1178 to remotely monitor vehicles.)
Yu does not explicitly disclose, however Ratliff further teaches: geometric fabric ([col. 5, lines 3-5] a computer system can be provided that generates a policy or geometric fabric that includes policy layers.)
As previously stated, Yu and Ratliff are analogous art to the claimed invention since they are from the similar field of neural network processing for pick and place robotic control. It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention, with a reasonable expectation for success, to modify the neural processing of Yu to enable the specific requirements for the network as described in Ratliff.
The motivation for modification would have been to provide the neural network processing disclosed in Yu with the particular processor taught in Ratliff for the purpose of improving training with a specific class of controller.
Regarding claim 11: The combination of Yu and Ratliff teaches: The one or more processors of claim 1,
Yu further discloses: wherein the one or more processors are comprised in at least one of: a control system for an autonomous or semi-autonomous machine; a perception system for an autonomous or semi-autonomous machine; a system for performing simulation operations; a system for performing digital twin operations; a system for performing light transport simulation; a system for performing collaborative content creation for 3D assets; a system for performing deep learning operations; a system implemented using an edge device; a system implemented using a robot; a system for performing conversational Al operations; a system for performing generative Al operations using a multi-modal language model; a system for performing generative Al operations using a large language model (LLM); a system for performing generative Al operations using a video language model (VLM); a system for generating synthetic data; a system incorporating one or more virtual machines (VMs); a system implemented at least partially in a data center; or a system implemented at least partially using cloud computing resources. ([0061] a processor comprises one or more circuits to use one or more neural networks to segment one or more objects of one or more images 102, 104 based, at least in part, on one or more bounding boxes identified by said one or more neural networks. In at least one embodiment, said processor performs such operations as described in FIG. 2, which shows one or more training processes for a framework 106 for object detection 108, instance segmentation 110, and semantic correspondence 112. [0062] In at least one embodiment, a first image 102 and/or a second image 104 are images captured from one or more images and/or video capturing devices, such as one or more systems of a vehicle (e.g., autonomous vehicle, semi-autonomous vehicle) or other imaging system (e.g., robotic device, grasping device, medical imaging device, satellite imaging system).)
Claim 7 is rejected under 35 U.S.C. 103 as being unpatentable over Yu et al. (US20220261593, referred to as Yu) in view of Ratliff et al. (US12240112, referred to as Ratliff) and further in view of Rhoads et al. (US10922957B2, referred to as Rhoads)
Regarding claim 7: The combination of Yu and Ratliff teaches: The one or more processors of claim 1,
Yu further discloses: wherein the one or more circuits are to: execute the simulation [at a frequency of about 120 Hertz.] ([0251] server(s) 1178 may be used to train machine learning models (e.g., neural networks) based at least in part on training data. In at least one embodiment, training data may be generated by vehicles, and/or may be generated in a simulation (e.g., using a game engine). In at least one embodiment, any amount of training data is tagged (e.g., where associated neural network benefits from supervised learning) and/or undergoes other pre-processing. In at least one embodiment, any amount of training data is not tagged and/or pre-processed (e.g., where associated neural network does not require supervised learning). In at least one embodiment, once machine learning models are trained, machine learning models may be used by vehicles (e.g., transmitted to vehicles over network(s) 1190), and/or machine learning models may be used by server(s) 1178 to remotely monitor vehicles.)
Yu does not disclose the following limitations, however Rhoads, from an analogous field of endeavor teaches: at a frequency of about 120 hz ([0040] the apparatus for providing mixed reality content may provide mixed reality content by executing the simulation pipeline 120 to update the entire virtual scene 130 and display it on the display in accordance with its own simulation cycle, i.e., a display refresh cycle (60 to 120 Hz).)
Yu, Ratliff, and Rhoads are analogous art to the claimed invention since they are from the similar field of neural network processing for pick and place robotic control. It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention, with a reasonable expectation for success, to modify the neural processing of Yu to enable the specific frequency taught in Lee.
The motivation for modification would have been to provide the neural network processing disclosed in Yu with the specific simulation frequency taught in Rhoads for the purpose of improving training with a common encoding processing frequency.
Claim 9 is rejected under 35 U.S.C. 103 as being unpatentable over Yu et al. (US20220261593, referred to as Yu) in view of Ratliff et al. (US12240112, referred to as Ratliff) and further in view of Kim et al. (US20250218164, referred to as Kim)
Regarding claim 9: The combination of Yu and Ratliff teaches: The one or more processors of claim 1,
Yu discloses: wherein the one or more circuits are to: modify [lighting or materials] of the simulation during the second update stage. ([0251] server(s) 1178 may be used to train machine learning models (e.g., neural networks) based at least in part on training data. In at least one embodiment, training data may be generated by vehicles, and/or may be generated in a simulation (e.g., using a game engine). In at least one embodiment, any amount of training data is tagged (e.g., where associated neural network benefits from supervised learning) and/or undergoes other pre-processing. In at least one embodiment, any amount of training data is not tagged and/or pre-processed (e.g., where associated neural network does not require supervised learning). In at least one embodiment, once machine learning models are trained, machine learning models may be used by vehicles (e.g., transmitted to vehicles over network(s) 1190), and/or machine learning models may be used by server(s) 1178 to remotely monitor vehicles.)
Yu does not explicitly disclose: modify lighting or materials of the simulation
Yu does not disclose the following limitations, however Kim, from an analogous field of endeavor, further teaches: modify lighting or materials of the simulation ([0140] the virtual space generation module 440 may change lighting or texture rendering for the virtual space in various ways before executing a simulation. As described above, the virtual space generation module 440 may adjust the illuminance of the virtual space based on the illuminance characteristic information about the target space.)
Yu, Ratliff, and Kim are analogous art to the claimed invention since they are from the similar field of neural network simulations of robots. It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention, with a reasonable expectation for success, to modify the neural processing of Yu to enable the update stage to include adjustments to the lighting of the simulation as described in Kim.
The motivation for modification would have been to provide the neural network processing disclosed in Yu with the particular lighting updating taught in Kim for the purpose of improving training over time by parametrization of the simulation lighting.
Claims 12-20 are rejected under 35 U.S.C. 103 as being unpatentable over Yu et al. (US20220261593, referred to as Yu) in view of Ratliff et al. (US12240112, referred to as Ratliff) and further in view of Jiao et al. (US20230004588, referred to as Jiao) and further in view of Lee et al. (US20250371748, referred to as Lee).
Regarding claim 12: Yu discloses: A system, comprising: an autonomous machine to operate in response to control instructions from a [geometric fabric] controller; and one or more processors to: capture at least two color-based images of an environment including the autonomous machine and a physical object; provide the at least two [color-based] images as input to a machine-learning model comprising an encoder to implement [cross-attention] masking between the at least two color-based images, ([0061] a processor comprises one or more circuits to use one or more neural networks to segment one or more objects of one or more images 102, 104 based, at least in part, on one or more bounding boxes identified by said one or more neural networks. In at least one embodiment, said processor performs such operations as described in FIG. 2, which shows one or more training processes for a framework 106 for object detection 108, instance segmentation 110, and semantic correspondence 112. [0062] In at least one embodiment, a first image 102 and/or a second image 104 are images captured from one or more images and/or video capturing devices, such as one or more systems of a vehicle (e.g., autonomous vehicle, semi-autonomous vehicle) or other imaging system (e.g., robotic device, grasping device, medical imaging device, satellite imaging system).) the machine-learning model generating at least one action for the autonomous machine; and ([0136] training framework 904 trains untrained neural network 906 repeatedly while adjust weights to refine an output of untrained neural network 906 using a loss function and adjustment algorithm, such as stochastic gradient descent. In at least one embodiment, training framework 904 trains untrained neural network 906 until untrained neural network 906 achieves a desired accuracy. In at least one embodiment, trained neural network 908 can then be deployed to implement any number of machine learning operations.) control the autonomous machine with respect to the physical object using the at least one action and a [geometric fabric] controller. ([0061] a processor comprises one or more circuits to use one or more neural networks to segment one or more objects of one or more images 102, 104 based, at least in part, on one or more bounding boxes identified by said one or more neural networks. In at least one embodiment, said processor performs such operations as described in FIG. 2, which shows one or more training processes for a framework 106 for object detection 108, instance segmentation 110, and semantic correspondence 112. [0062] In at least one embodiment, a first image 102 and/or a second image 104 are images captured from one or more images and/or video capturing devices, such as one or more systems of a vehicle (e.g., autonomous vehicle, semi-autonomous vehicle) or other imaging system (e.g., robotic device, grasping device, medical imaging device, satellite imaging system).)
Yu does not explicitly disclose: geometric fabric … cross-attention … color-based
Yu does not disclose the following limitations, however Ratliff, in an analogous field of endeavor teaches: geometric fabric … ([col. 5, lines 3-5] a computer system can be provided that generates a policy or geometric fabric that includes policy layers.)
As previously stated, Yu and Ratliff are analogous art to the claimed invention since they are from the similar field of neural network processing for pick and place robotic control. It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention, with a reasonable expectation for success, to modify the neural processing of Yu to enable the specific requirements for the imaging and network as described in Ratliff.
The motivation for modification would have been to provide the neural network processing disclosed in Yu with the particular processor taught in Ratliff for the purpose of improving training with a specific class of controller.
Yu does not explicitly disclose: cross-attention
Yu does not disclose the following limitations, however Jiao, from an analogous field of endeavor, further teaches: cross-attention ([0083] the encoder-decoder attention mechanism 1008 performs cross-attention based on the output information generated by the encoder 804 and the output information supplied by the preceding component in the decoder block 1002)
Yu, Ratliff, and Jiao are analogous art to the claimed invention since they are from the similar field of neural network processing for pick and place robotic control. It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention, with a reasonable expectation for success, to modify the neural processing of Yu to enable the common encoder output cross-attention taught in Jiao.
The motivation for modification would have been to provide the neural network processing disclosed in Yu with the cross-attention taught in Jiao for the purpose of improving training with a common encoding processing technique.
Yu does not explicitly disclose: color-based
Yu does not disclose the following limitations, however Sawada, from an analogous field of endeavor, further teaches: color-based ([0063] this method may include an image penalty calculating step for calculating the correction amount of the input color image based on the image penalty (here a penalty function for defining the unnaturalness of this color image) to be defined as to the color image input in the image simulation step, and a feedback step for correcting the color image input in the image simulating step based on above-mentioned expanded comparison result and the correction amount.)
Yu, Ratliff, Jiao, and Lee are analogous art to the claimed invention since they are from the similar field of neural network image processing. It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention, with a reasonable expectation for success, to modify the neural processing of Yu to enable the common color image input taught in Sawada.
The motivation for modification would have been to provide the neural network processing disclosed in Yu with the color-based processing taught in Sawada.
Regarding claim 13: The combination of Yu, Ratliff, Jiao, and Sawada teaches: The system of claim 12,
Yu further discloses: wherein the machine-learning model is to generate a predicted position of the object, and wherein the one or more processors are to control the autonomous machine further based at least on the predicted position of the object. ([0061] a processor comprises one or more circuits to use one or more neural networks to segment one or more objects of one or more images 102, 104 based, at least in part, on one or more bounding boxes identified by said one or more neural networks. In at least one embodiment, said processor performs such operations as described in FIG. 2, which shows one or more training processes for a framework 106 for object detection 108, instance segmentation 110, and semantic correspondence 112. [0062] In at least one embodiment, a first image 102 and/or a second image 104 are images captured from one or more images and/or video capturing devices, such as one or more systems of a vehicle (e.g., autonomous vehicle, semi-autonomous vehicle) or other imaging system (e.g., robotic device, grasping device, medical imaging device, satellite imaging system).)
Regarding claim 14: The combination of Yu, Ratliff, Jiao, and Sawada teaches: The system of claim 12,
Yu further discloses: wherein the one or more processors are to control the autonomous machine further based at least on an output of a state machine. ([0061] a processor comprises one or more circuits to use one or more neural networks to segment one or more objects of one or more images 102, 104 based, at least in part, on one or more bounding boxes identified by said one or more neural networks. In at least one embodiment, said processor performs such operations as described in FIG. 2, which shows one or more training processes for a framework 106 for object detection 108, instance segmentation 110, and semantic correspondence 112. [0062] In at least one embodiment, a first image 102 and/or a second image 104 are images captured from one or more images and/or video capturing devices, such as one or more systems of a vehicle (e.g., autonomous vehicle, semi-autonomous vehicle) or other imaging system (e.g., robotic device, grasping device, medical imaging device, satellite imaging system).)
Regarding claim 15: The combination of Yu, Ratliff, Jiao, and Sawada teaches: The system of claim 12,
Yu further discloses: wherein the one or more processors are to: provide a set of proprioception data and the at least two color-based images as input to the machine-learning model to generate the at least one action. ([0061] a processor comprises one or more circuits to use one or more neural networks to segment one or more objects of one or more images 102, 104 based, at least in part, on one or more bounding boxes identified by said one or more neural networks. In at least one embodiment, said processor performs such operations as described in FIG. 2, which shows one or more training processes for a framework 106 for object detection 108, instance segmentation 110, and semantic correspondence 112. [0062] In at least one embodiment, a first image 102 and/or a second image 104 are images captured from one or more images and/or video capturing devices, such as one or more systems of a vehicle (e.g., autonomous vehicle, semi-autonomous vehicle) or other imaging system (e.g., robotic device, grasping device, medical imaging device, satellite imaging system).)
Regarding claim 16: Rejected using the same rationale as claim 2.
Regarding claim 17: Rejected using the same rationale as claim 11.
Regarding claim 18: Rejected using the same rationale as claim 1.
Regarding claim 19: Rejected using the same rationale as claim 2 and 16.
Regarding claim 20: Rejected using the same rationale as claim 3.
Conclusion
The prior art made of record, and not relied upon, considered pertinent to applicant' s disclosure or directed to the state of art is listed on the enclosed PTO-892.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ATTICUS A CAMERON whose telephone number is 703-756-4535. The examiner can normally be reached M-F 8:30 am - 4:30 pm.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Thomas Worden can be reached on 571-272-4876. 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.
/ATTICUS A CAMERON/ /JASON HOLLOWAY/ Primary Examiner, Art Unit 3658
Examiner, Art Unit 3658A
Prosecution Insights