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 .
Response to Remark(s)
Applicant's amendment filed January 21st, 2026 has been fully entered and considered. Applicant’s amendment to the claims have overcome each and every 112b rejections previously set forth in the Non-Final Office Action mailed on October 21st, 2025. Regarding the arguments to the previous prior art rejections, the examiner respectfully finds the arguments to be non-persuasive, see response to remarks section below. Accordingly, this action is made final.
Status of Claims
Claims 1-7 are pending, claim 3 has been amended, claim 8 has been canceled. Claims 1-7 remains rejected.
Response to Argument(s)
112b rejection:
The Applicants amended the dependent claim 3 to overcome the 112b rejection, argument found to be persuasive, the 112b rejection is withdrawn.
102 and 103 rejections:
In pages 6-8 of the remarks, the Applicants argue that the proposed prior arts Klein, Paul, Tripathi, and Penna, alone or in combination fail to disclose or suggest each and every limitation of claim 1, such as particularly for the limitation of:
“perform compression encoding on each of a first image obtained from an image generation apparatus and a second image indicating a detection target object so as to extract a feature quantity that allows object detection and so as to be decoded later, thereby generate respective one of first encoding information that is usable as a first feature quantity of the first image and that is compressed, and second encoding information that is usable as a second feature quantity of the second image and that is compressed”
In support of the above argument, the Applicants assert that the proposed Klein (previously used in the 102 rejection of the independent claim 1) discloses not using decompression in the context of feature vector such as, “unlike traditional feature vectors compression which decompresses before applying pairwise matching, the current suggestion omits the decompression stage, and perform pairwise matching directly on the compressed data,” Klein’s pp. 140. Therefore, Klein cannot perform the recited “compression encoding…so as to extract a feature quantity that allows object detection and so as to be decided later” as recited in claim 1.
The applicants further state that other prior arts do not remedy this deficiency and are not relied on by the examiner to do so.
Examiner’s reply:
The examiner respectfully disagrees with the Applicants’ argument and find the argument to be incommensurate with the scope of the claims, importantly, the Applicants are reminded that the claims are construed based on BRI (broadest reasonable interpretation) in light of the specification, therefore, the examiner find the claim’s scope to fall under the teaching of Klein.
And that the claim, such as for the independent claim 1, recites nowhere in the claim a decompression step, the claim recites “perform compression….of the first image and that is compressed…..of the second image and that is compressed” which is not commensurate with the Applicants’ argument that stated the proposed Klein teaches “unlike traditional feature vectors compression which decompresses before applying pairwise matching, the current suggestion omits the decompression stage, and perform pairwise matching directly on the compressed data,” Klein’s pp. 140. Therefore, Klein cannot perform the recited “compression encoding…so as to extract a feature quantity that allows object detection and so as to be decided later” as recited in claim 1. Specifically, the claim recites a compression step, and the Applicants’ argument is mentioning Klein omits the decompression stage, which are two different terms, different concepts, and are not analogous. Importantly, the examiner finds Klein to teach the claim 1’s limitations, such as:
Klein teaches perform compression encoding on each of a first image obtained from an image generation apparatus (section 2, 1st 3 paragraphs, discloses compressing of query image and image from a database to have two SIFT feature vectors of the two images [as disclosed in section 1, last 3 paragraphs wherein the query image obtained feature vector being pairwise matched with an image’s feature vector from a database for object matching and object detection]; therefore, any of the two images can be understood to be the first image as claimed, such as the image from the database; the computing circuit used to obtain the feature vector for this image can be understood to be analogous to the claimed image generation apparatus as claimed, by BRI [broadest reasonable interpretation]) and a second image indicating a detection target object so as to extract a feature quantity that allows object detection (and the query image, as discussed previously, can be understood to be analogous to the second image as claimed which indicates the object to the query of the detection target object as claimed, its feature vector can be understood to be a feature quantity as claimed, by BRI) and so as to be decoded later (the compressed feature vectors as discussed previously, as disclosed in section 2, 1st 3 paragraphs, to be decoded later as disclosed in section 2, 4th par.), thereby generate respective one of first encoding information that is usable as a first feature quantity of the first image, image and that is compressed, and second encoding information that is usable as a second feature quantity of the second image and that is compressed (the compressed 2 SIFT feature vectors as discussed previously, as disclosed in section 2, 1st 3 pars., as for the query image [second image] and the image from the database [1st image], the compressed features vectors are analogous to the claimed first feature quantity and second feature quantity, respectively, as claimed, by BRI, both are compressed); and detect the detection target object in the first image, by using the first and second feature quantities (as perform the object pairwise matching or the object detection using the two compressed feature vectors or the first and the second feature quantities as claimed, by BRI, as disclosed in section 2).
See prior art rejections below for more details.
Therefore, the prior art rejections remain.
Claim Rejections - 35 USC § 102
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
Claims 1-3 and 6-7 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Shmuel T. Klein et. al. (“Metric Preserving Dense SIFT Compression, 2014, Proceedings of the Prague Stringology Conference 2014” hereinafter as “Klein”).
Regarding claim 1, Klein discloses an object detection apparatus comprising (section 1, 2nd par., discloses the invention is for object detection computing device): at least one memory configured to store instructions; and at least one processor configured to execute the instructions to (section 1, 3rd par., discloses a computing method hence can be understood to have the use of a computer, which includes memory storing instructions of the invention to be executed by a processor): perform compression encoding on each of a first image obtained from an image generation apparatus (section 2, 1st 3 paragraphs, discloses compressing of query image and image from a database to have two SIFT feature vectors of the two images [as disclosed in section 1, last 3 paragraphs wherein the query image obtained feature vector being pairwise matched with an image’s feature vector from a database for object matching and object detection]; therefore, any of the two images can be understood to be the first image as claimed, such as the image from the database; the computing circuit used to obtain the feature vector for this image can be understood to be analogous to the claimed image generation apparatus as claimed, by BRI [broadest reasonable interpretation]) and a second image indicating a detection target object so as to extract a feature quantity that allows object detection (and the query image, as discussed previously, can be understood to be analogous to the second image as claimed which indicates the object to the query of the detection target object as claimed, its feature vector can be understood to be a feature quantity as claimed, by BRI) and so as to be decoded later (the compressed feature vectors as discussed previously, as disclosed in section 2, 1st 3 paragraphs, to be decoded later as disclosed in section 2, 4th par.), thereby generate respective one of first encoding information that is usable as a first feature quantity of the first image, image and that is compressed, and second encoding information that is usable as a second feature quantity of the second image and that is compressed (the compressed 2 SIFT feature vectors as discussed previously, as disclosed in section 2, 1st 3 pars., as for the query image [second image] and the image from the database [1st image], the compressed features vectors are analogous to the claimed first feature quantity and second feature quantity, respectively, as claimed, by BRI, both are compressed); and detect the detection target object in the first image, by using the first and second feature quantities (as perform the object pairwise matching or the object detection using the two compressed feature vectors or the first and the second feature quantities as claimed, by BRI, as disclosed in section 2).
Regarding claim 2, Klein discloses the object detection apparatus according to claim 1, wherein the at least one processor is configured to execute the instructions to transmit the first encoding information to an information processing apparatus that performs a predetermined operation using the first encoding information (the processing of the feature vector mapped to be the first encoding information as discussed above in claim 1, which is being processed by a processor, hence the processing includes inputting the feature vector into another sequential processing therefore, by BRI, it can be understood to be analogous to transmitting the first encoding information to an information processing apparatus [the circuit part of the computer that is responsible for inputting the feature vector into the subsequential processing]), through a communication line (the data being transferred in the circuit include path wire or path connections, based on BRI, is analogous to through a communication line as claimed).
Regarding claim 3, Klein discloses the object detection apparatus according to claim 2, wherein the predetermined operation (as discussed above in claim 2) includes: a first operation of decoding the first encoding information, thereby to generate a third image; a second operation of analyzing the third image; a third operation of storing the first encoding information in a storage apparatus; and a fourth operation of storing the third image in a storage apparatus (the examiner selects “a third operation of storing the first encoding information in a storage apparatus” which is taught in Klein, as being mapped previously in claims 1 and 2, the feature vector being mapped to the recited first encoding information being processed by the computer hence, can be understood to be stored in the computer as well included in RAM or ROM storage of the computer [analogous to the recited storage apparatus as claimed, by BRI]).
Regarding claim 6, Klein discloses an object detection system comprising an object detection apparatus and an information processing apparatus, the object detection apparatus including: at least one first memory configured to store instructions; and at least one first processor configured to execute the instructions to (section 1, 2nd par., discloses the invention is for object detection computing device; section 1, 3rd par., discloses a computing method hence can be understood to have the use of a computer, which includes memory storing instructions of the invention to be executed by a processor): perform compression encoding on each of a first image obtained from an image generation apparatus (section 2, 1st 3 paragraphs, discloses compressing of query image and image from a database to have two SIFT feature vectors of the two images [as disclosed in section 1, last 3 paragraphs wherein the query image obtained feature vector being pairwise matched with an image’s feature vector from a database for object matching and object detection]; therefore, any of the two images can be understood to be the first image as claimed, such as the image from the database; the computing circuit used to obtain the feature vector for this image can be understood to be analogous to the claimed image generation apparatus as claimed, by BRI [broadest reasonable interpretation]) and a second image indicating a detection target object so as to extract a feature quantity that allows object detection (and the query image, as discussed previously, can be understood to be analogous to the second image as claimed which indicates the object to the query of the detection target object as claimed, its feature vector can be understood to be a feature quantity as claimed, by BRI) and so as to be decoded later (the compressed feature vectors as discussed previously, as disclosed in section 2, 1st 3 paragraphs, to be decoded later as disclosed in section 2, 4th par.), thereby generate respective one of first encoding information that is usable as a first feature quantity of the first image, image and that is compressed, and second encoding information that is usable as a second feature quantity of the second image and that is compressed (the compressed 2 SIFT feature vectors as discussed previously, as disclosed in section 2, 1st 3 pars., as for the query image [second image] and the image from the database [1st image], the compressed features vectors are analogous to the claimed first feature quantity and second feature quantity, respectively, as claimed, by BRI, both are compressed); and detect the detection target object in the first image, by using the first and second feature quantities (as perform the object pairwise matching or the object detection using the two compressed feature vectors or the first and the second feature quantities as claimed, by BRI, as disclosed in section 2); transmit the first encoding information to the information processing apparatus (the processing of the feature vector mapped to be the first encoding information as discussed above in claim 1, which is being processed by a processor, hence the processing includes inputting the feature vector into another sequential processing therefore, by BRI, it can be understood to be analogous to transmitting the first encoding information to an information processing apparatus [the circuit part of the computer that is responsible for inputting the feature vector into the subsequential processing]) , through a communication line (the data being transferred in the circuit include path wire or path connections, based on BRI, is analogous to through a communication line as claimed), the information processing apparatus including: at least one second memory configured to store instructions (the circuit part to perform these specific steps of the invention, as discussed previously, to be mapped to the recited information processing apparatus; therefore, the memory part that store the program to perform these specific functions can be understood to be analogous to the second memory as claimed, by BRI); and at least one second processor configured to execute the instructions to perform a predetermined operation using the first encoding information (and the processor part of the circuit that is specialty preprogramed to perform these specific steps is analogous to the recited second processor as claimed, by BRI, which is to perform the operation as claimed, by BRI).
Regarding claim 7, Klein discloses an object detection method comprising (section 1, 2nd par., discloses the invention is for object detection computing): performing compression coding on each of a first image obtained from an image generation apparatus (section 2, 1st 3 paragraphs, discloses compressing of query image and image from a database to have two SIFT feature vectors of the two images [as disclosed in section 1, last 3 paragraphs wherein the query image obtained feature vector being pairwise matched with an image’s feature vector from a database for object matching and object detection]; therefore, any of the two images can be understood to be the first image as claimed, such as the image from the database; the computing circuit used to obtain the feature vector for this image can be understood to be analogous to the claimed image generation apparatus as claimed, by BRI [broadest reasonable interpretation]) and a second image indicating a detection target object so as to extract a feature quantity that allows object detection (and the query image, as discussed previously, can be understood to be analogous to the second image as claimed which indicates the object to the query of the detection target object as claimed, its feature vector can be understood to be a feature quantity as claimed, by BRI) and so as to be decoded later (the compressed feature vectors as discussed previously, as disclosed in section 2, 1st 3 paragraphs, to be decoded later as disclosed in section 2, 4th par.), thereby generating respective one of first encoding information that is usable as a first feature quantity of the first image, image and that is compressed, and second encoding information that is usable as a second feature quantity of the second image and that is compressed (the compressed 2 SIFT feature vectors as discussed previously, as disclosed in section 2, 1st 3 pars., as for the query image [second image] and the image from the database [1st image], the compressed features vectors are analogous to the claimed first feature quantity and second feature quantity, respectively, as claimed, by BRI, both are compressed); and detecting the detection target object in the first image, by using the first and second feature quantities (as perform the object pairwise matching or the object detection using the two compressed feature vectors or the first and the second feature quantities as claimed, by BRI, as disclosed in section 2).
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 4-5 are rejected under 35 U.S.C. 103 as being unpatentable over Shmuel T. Klein et. al. (“Metric Preserving Dense SIFT Compression, 2014, Proceedings of the Prague Stringology Conference 2014” hereinafter as “Klein”) in view of Sayan Paul et. al. (“Software development using context aware searching of components in large repositories, May 2015, International Conference on Computing, Communication & Automation” hereinafter as “Paul”) and further in view of Subarna Tripathi et. al. (“Context Matters: Refining Object Detection in Video with Recurrent Neural Networks, Jul. 2016, arXiv: 1607.04648v2” hereinafter as “Tripathi”) and Barbara Penna et. al. (“Transform Coding Techniques for Lossy Hyperspectral Data Compression, May 2007, IEEE Transactions on Geoscience and Remote Sensing, Vol. 45, Issue 5” hereinafter as “Penna”).
Regarding claim 4, Klein discloses the object detection apparatus according to claim 1, wherein the at least one processor is configured to execute the instructions to (as discussed above in claim 1): generate the first and second encoding information that are respectively usable as the first and second feature quantities (as discussed above in claim 1, the compressed 2 SIFT feature vectors as discussed previously, as disclosed in section 2, 1st 3 pars., as for the query image and the image from the database, the compressed features vectors are analogous to the claimed first feature quantity and second feature quantity, respectively, as claimed, by BRI, both are compressed; and therefore, the feature vector of the query image include second encoding information and the feature vector of the image from the database include the first encoding information), by using a first model part that outputs the first and second encoding information when the first and second images are inputted (the model used for compressing the feature vectors using Fibonacci Code as disclosed in section 2, 1st 2 pars., hence, the Fibonacci Code can be understood to be analogous to the recited first model part as claimed, by BRI); and detect the detection target object, by using a second model part that outputs a detection result of the detection target object in the first image when the first and second feature quantities are inputted (the object matching as discussed above in claim 1 which is used for object detection, the object detection as discussed is done by matching of the object in the query image with the images in the database through the feature vectors as discussed previously as disclosed in section 4; moreover, the object detection is done by using the compressed feature vectors and the object detection is done using the algorithm in section 4 [second model part as claimed, by BRI]), of the computational model (of the computing of the invention of Klein of the computing model, by BRI, analogous to the recited computational model).
However, Klein does not explicitly disclose the first model part of a computational model generated by machine learning; and the computational model is generated by machine learning using a first loss function and a second loss function, the first loss function being based on an error between the detection result of the detection target object outputted by the second model part of the computational model to which a fourth image for learning is inputted and a ground truth label of the detection result of the detection target object in the fourth image, the second loss function being based on an error between a third image generated by decoding the first encoding information outputted by the first model part of the computational model to which the fourth image is inputted and the fourth image.
In the same field of using Fibonacci code for querying of feature (FIG. 1 and page 767, last par., Paul) Paul discloses the first model part of a computational model generated by machine learning (the Fibonacci Code of page 767, last par., as part of a neural network for querying features; by BRI, covers the scope of the claimed feature).
Thus, it would have been obvious for a person of ordinary skill in the art before the effective filing date to modify Klein to perform generating the first and second encoding information that are respectively usable as the first and second feature quantities by using a first model part that outputs the first and second encoding information when the first and second images are inputted; moreover, the first model part of a computational model generated by machine learning as taught by Paul to arrive at the claimed invention discussed above. Such a modification is the result of combing prior art elements according to known methods to yield predictable results. The motivation for the proposed modification would have been to perform querying using the method as discussed more effectively (abstract, Paul).
However, Klein in view of Paul does not explicitly disclose the computational model is generated by machine learning using a first loss function and a second loss function, the first loss function being based on an error between the detection result of the detection target object outputted by the second model part of the computational model to which a fourth image for learning is inputted and a ground truth label of the detection result of the detection target object in the fourth image, the second loss function being based on an error between a third image generated by decoding the first encoding information outputted by the first model part of the computational model to which the fourth image is inputted and the fourth image.
In the same field of object detection (title, Tripathi), Tripathi discloses the computational model is generated by machine learning using a first loss function and a second loss function (Tripathi discloses using neural network for object detection using a loss function such a disclosed in section 2.1.1, last par., which can be understood to be analogous to the recited first loss function as claimed; moreover, the loss function includes other loss value such as shown in equation which, any can be understood to be the second loss function as claimed, by BRI), the first loss function being based on an error between the detection result of the detection target object outputted by the second model part of the computational model to which a fourth image for learning is inputted and a ground truth label of the detection result of the detection target object in the fourth image (the loss function of section 2.1.1 is based on ground truth and prediction of the object of a bounding box predictor [of the ground truth would be understood to be analogous to the recited ground truth label as claimed, by BRI]; the ground truth image here is analogous to the recited fourth image as claimed, by BRI, which is being input into the loss function together with the prediction to find the loss output, the loss function here is for error between the ground truth and the prediction; by BRI, covers the scope of the claimed limitation).
Thus, it would have been obvious for a person of ordinary skill in the art before the effective filing date to modify Klein in view of Paul to perform generating the first and second encoding information that are respectively usable as the first and second feature quantities by using a first model part that outputs the first and second encoding information when the first and second images are inputted; moreover, the first model part of a computational model generated by machine learning and the computational model is generated by machine learning using a first loss function and a second loss function, the first loss function being based on an error between the detection result of the detection target object outputted by the second model part of the computational model to which a fourth image for learning is inputted and a ground truth label of the detection result of the detection target object in the fourth image as taught by Tripathi to arrive at the claimed invention discussed above. Such a modification is the result of combing prior art elements according to known methods to yield predictable results. The motivation for the proposed modification would have been to predict object more accurately using the method as discussed (abstract, Tripathi).
However, Klein in view of Paul and Tripathi does not explicitly disclose the second loss function being based on an error between a third image generated by decoding the first encoding information outputted by the first model part of the computational model to which the fourth image is inputted and the fourth image.
In the same field of compression encoding of images (abstract, Penna) Penna discloses the second loss function being based on an error between a third image generated by decoding the first encoding information outputted by the first model part of the computational model to which the fourth image is inputted and the fourth image (page 1417, last 3 paragraphs, discloses computing error for the detector run on the decoded images [analogous to the first encoding information as claimed, by BRI] and ground truth [analogous to the fourth image as claimed], also as disclosed in page 1418, section B, 2nd to the last paragraph; moreover, the error here is used for performance validation according to page 1418, section B, last par., as a metric to determine performance loss such as disclosed in page 1414, section B, and page 1415, 1st par.; by BRI, it is analogous to the claimed limitation as the ground truth and the decoded feature vector are being inputted into the equation for calculation of the error, moreover, the decoded feature vector here include image data can be understood to be analogous to the third image as claimed, by BRI, and here the performance loss can be understood to be analogous to the second loss function as claimed, by BRI).
Thus, it would have been obvious for a person of ordinary skill in the art before the effective filing date to modify Klein in view of Paul and Tripathi to perform generating the first and second encoding information that are respectively usable as the first and second feature quantities by using a first model part that outputs the first and second encoding information when the first and second images are inputted; moreover, the first model part of a computational model generated by machine learning and the computational model is generated by machine learning using a first loss function and a second loss function, the first loss function being based on an error between the detection result of the detection target object outputted by the second model part of the computational model to which a fourth image for learning is inputted and a ground truth label of the detection result of the detection target object in the fourth image, and the second loss function being based on an error between a third image generated by decoding the first encoding information outputted by the first model part of the computational model to which the fourth image is inputted and the fourth image as taught by Penna to arrive at the claimed invention discussed above. Such a modification is the result of combing prior art elements according to known methods to yield predictable results. The motivation for the proposed modification would have been to perform detection more accurately (abstract, Penna).
Regarding claim 5, Klein in view of Paul further in view of Tripathi and Penna discloses the object detection apparatus according to claim 4 (as discussed above in claim 4), wherein the computational model includes a neural network (as discussed above in claim 4, the computation model includes a neural network), and the first model part includes an encoder part of an autoencoder (as taught in Klein, the model [which was mapped to the recited first model part] for encoding the feature vector includes an encoder such as disclosed in page 140, 2nd par of Klein; moreover, the system of Klein includes an encoder to compress the input data and use decoder to reconstruct the data such as disclosed above in claim 1 therefore, it’s analogous to a system of an autoencoder as claimed, by BRI).
Conclusion
THIS ACTION IS MADE FINAL. 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 nonprovisional extension fee (37 CFR 1.17(a)) 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 mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to PHUONG HAU CAI whose telephone number is (571)272-9424. The examiner can normally be reached M-F 8:30 am - 5:00pm.
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, Chineyere Wills-Burns can be reached 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.
/PHUONG HAU CAI/Examiner, Art Unit 2673
/CHINEYERE WILLS-BURNS/Supervisory Patent Examiner, Art Unit 2673
Prosecution Insights