CTNF 18/793,309 CTNF 100467 Notice of Pre-AIA or AIA Status 07-03-aia AIA 15-10-aia The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA. Claim Rejections - 35 USC § 112 Claim 8 is rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention. It is unclear from the claim language whether the term “register” refers to the image processing operation of “registration” or “registering” as a task for saving information about the identity of a subject in an image in a data structure. For examination purposes, this limitation is being interpreted as the second definition (saving identifying image data about a plurality of persons). Claim Rejections - 35 USC § 102 07-06 AIA 15-10-15 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. 07-07-aia AIA 07-07 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 – 07-08-aia AIA (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. 07-12-aia AIA (a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention. 07-15 AIA Claim s 1-8 and 11-12 are rejected under 35 U.S.C. 102( a)(1 ) as being anticipated by Yamamoto et al. (Japanese PG Pub 2022160064, hereinafter “Yamamoto”) . Regarding claim 1, Yamamoto describes a method for lowering the processing load of a facial detection and recognition system, said method including an information processing apparatus comprising: one or more memories storing instructions (para. 0068, “For example, the information processing apparatus 100 has at least a CPU 1511, a ROM 1512, a RAM 1513, an auxiliary storage device 1514, a display section 1515, an operation section 1516, a communication I/F 1517, and a bus 1518 as a hardware configuration. The CPU 1511 controls the entire information processing apparatus 100 using computer programs and data stored in the ROM 1512 and the RAM 1513, and executes information processing including the aforementioned face recognition processing, recognition target range determination processing, and the like”); and one or more processors (para. 0068, “For example, the information processing apparatus 100 has at least a CPU 1511…The CPU 1511 controls the entire information processing apparatus 100 using computer programs and data stored in the ROM 1512 and the RAM 1513, and executes information processing including the aforementioned face recognition processing, recognition target range determination processing, and the like.”) that, upon execution of the stored instructions, are configured to: extract feature information from an input image (para. 0019, “As an example, the face recognition unit 103 performs feature extraction processing (extraction of humanness) on both the face region image detected by the face detection unit 102 and the registered face image, A method or the like of calculating the distance between feature vectors obtained as a result is used. For the feature extraction processing, for example, a technique using a deep net can be used, and the Euclidean distance or cosine distance can be used as the distance between feature vectors.”); in a case where it is determined that the extracted feature information is inappropriate for object recognition, accumulate the feature information (para. 0013 for the determination of inappropriacy of the extracted feature information, “Hereinafter, this face detection condition will be referred to as "inappropriate detection condition". That is, in the face detection process, a face image determined to satisfy the unsuitable detection condition is determined to be an unregistered person in the recognition process, regardless of whether the person in the face image is a registered person.”; and paras. 0013 and 0024 for the accumulation, “In the present embodiment, in order to specify the unsuitable detection condition (portion with poor image quality), it is recommended to accumulate and manage the correspondence between the information (for example, the position in the image) obtained at the time of face detection and the similarity score” (0013); and “The accumulation unit 502 accumulates score correspondence data associated by the correspondence management unit 501 .FIG. 6 is a diagram showing an example of score correspondence data accumulated in the accumulation unit 502. As shown in FIG. As shown in FIG. 6, the accumulation unit 502 accumulates, as score correspondence data, data in which a management number, horizontal and vertical coordinates in a detected face image, and a maximum similarity score are associated with each other. It is Data of management numbers 1 to 4 shown in FIG. 6 correspond to the face image areas obtained from the face areas 400 to 403 in FIG. 4, respectively. For example, in the data of management number 1, the value of the maximum similarity score calculated by the face recognition processing is 300 for the face area indicated by the horizontal and vertical coordinates (200, 100) in the image. is shown.” (0024)); and based on feature information extracted from an image as a processing target and the accumulated feature information, determine whether the image as the processing target is an inappropriate image for object recognition (para. 0030, “Then, when there is a predetermined number of samples or more of threshold-exceeding data in the accumulated data output from the accumulating unit 502, the range determining unit 503 extracts coordinates from the threshold-exceeding data, and extracts coordinates from the threshold-exceeding data. identify.For example, when score correspondence data such as that shown in FIG. Extract. Then, the range determination unit 503 defines coordinates (400, 200), coordinates (400, 300), coordinates (500, 300), and coordinates (500, 200) as the vertices of a circumscribing rectangle that includes these coordinates. Specifies the rectangle to Hereinafter, the area inside the rectangle will be referred to as the "recognition target range", and the area outside the rectangle will be referred to as the "non-recognition target range". In this embodiment, the "non-recognition target range" corresponds to the "inappropriate detection condition"”, wherein the method and system disclosed by Yamamoto identifies a particular range within the photo as a suitable target (the “recognition target range”), while the unsuitable range outside of the range corresponds to the “inappropriate detection condition” for object recognition). Claims 11 and 12 are rejected, mutatis mutandis , for reasons similar to claim 1. Yamamoto further discloses the non-transitory computer-readable storage medium of claim 12 (para. 0068, “For example, the information processing apparatus 100 has at least a CPU 1511, a ROM 1512, a RAM 1513, an auxiliary storage device 1514, a display section 1515, an operation section 1516, a communication I/F 1517, and a bus 1518 as a hardware configuration. The CPU 1511 controls the entire information processing apparatus 100 using computer programs and data stored in the ROM 1512 and the RAM 1513, and executes information processing including the aforementioned face recognition processing, recognition target range determination processing, and the like”). Regarding claim 2, Yamamoto discloses all limitations of claim 1. Yamamoto further discloses wherein in a case where the input image is at least one of a blurred image and an incorrect detection image obtained by incorrectly detecting a target object, it is determined that the feature information regarding the input image is inappropriate for object recognition (para. 0013, “Hereinafter, this face detection condition will be referred to as "inappropriate detection condition". That is, in the face detection process, a face image determined to satisfy the unsuitable detection condition is determined to be an unregistered person in the recognition process, regardless of whether the person in the face image is a registered person. Therefore, it is possible to determine that a face image determined to satisfy the unsuitability detection condition is an unregistered person without performing recognition processing. For example, when a face is detected in a portion with poor image quality due to lens distortion, etc., there is a high possibility that the original facial features cannot be extracted well, so processing efficiency can be improved by omitting recognition processing. In the present embodiment, in order to specify the unsuitable detection condition (portion with poor image quality), it is recommended to accumulate and manage the correspondence between the information (for example, the position in the image) obtained at the time of face detection and the similarity score”, wherein the low-quality image is the blurred image, and wherein the image is deemed inappropriate for recognition as a result). Regarding claim 3, Yamamoto discloses all limitations of claim 1. Yamamoto further discloses wherein based on a similarity between the feature information extracted from the image as the processing target and the accumulated feature information, it is determined whether the image as the processing target is an inappropriate image (paras. 0012-0013, “On the other hand, when the maximum similarity score is smaller than the predetermined threshold, it is determined that the person detected from the input image is not one of the registered persons (determined as an unregistered person). Determination as a non-registered person is made "when the person in the input image is not really any of the registered persons". In addition to this, the judgment of a non-registered person is based on the fact that the person in the input image is a registered person because the shooting conditions of the input image were not good and a face image suitable for face recognition processing could not be obtained…Hereinafter, this face detection condition will be referred to as "inappropriate detection condition"”). Regarding claim 4, Yamamoto discloses all limitations of claim 1. Yamamoto further discloses wherein in a case where it is determined that the feature information extracted from the input image is inappropriate, the input image is accumulated (para. 0013, “In the present embodiment, in order to specify the unsuitable detection condition (portion with poor image quality), it is recommended to accumulate and manage the correspondence between the information (for example, the position in the image) obtained at the time of face detection and the similarity score”; and para. 0028, “The accumulation unit 502 sequentially adds and accumulates the score corresponding data in this way. Hereinafter, the score corresponding data accumulated by the accumulation unit 502 will be referred to as "accumulated data". Note that FIG. 6 shows an example of accumulated data in the case where one person passes through a corridor while being photographed by the surveillance camera 200 . As the number of people photographed by surveillance camera 200 increases, the accumulated data (score corresponding data) shown in FIG. 6 also increases. Then, the accumulated data of the accumulation unit 502 is output to the range determination unit 503 .”). Regarding claim 5, Yamamoto discloses all limitations of claim 1. Yamamoto further discloses wherein the one or more processors are further configured to detect a target object in the input image and extract an object area image (para. 0011, “In recognition processing for recognizing a target object, detection processing for detecting the target object is performed before performing target recognition processing. When performing face recognition as in the example of this embodiment, face detection processing is first performed to identify the position of a face image from an input image. Thereafter, recognition processing (face recognition processing) for identifying an individual is performed on the detected face image”, wherein the target object is the face and the object area image is the region-of-interest face image). Regarding claim 6, Yamamoto discloses all limitations of claim 1. Yamamoto further discloses wherein a similarity between pieces of feature information regarding two object area images is acquired (para. 0011, “In face recognition processing, a face image detected from an input image is compared with a pre-registered face image to calculate a recognition score (similarity score) representing the degree of similarity between the face images. Then, when the maximum similarity score among the similarity scores calculated for each registered person is greater than a predetermined threshold, the person ID (identification information) of the registered image with the maximum similarity score is input. The person detected from the image is specified as the person ID.”), wherein the acquired similarity is adjusted (para. 0036, “Further, the information processing apparatus 100 of the present embodiment collects and accumulates score correspondence data in order to specify the "inappropriate detection condition". Then, based on the collected score correspondence data, the information processing apparatus 100 specifies shooting conditions such that the similarity score does not exceed the recognition threshold even if the target person is a registered person. In other words, the information processing apparatus 100 selects an area in which the target person's face is too small in the image because it is far from the monitoring camera 200, an area in which the target person is in a position close to the monitoring camera 200, and therefore has a large downward angle. , a region whose similarity score does not exceed the threshold”, Examiner notes that this adjustment process of only adjusting the similarity score when an inappropriate image is detected and maintaining the similarity score if both images are appropriate is representative of the similarity adjustment unit 108 of the instant application, see at least para. 0049. In that situation, the adjustment is essentially multiplying the obtained similarity score by 1 ), and wherein based on the adjusted similarity, it is determined whether target objects present in the two object area images are the same as each other (para. 0020, “In this manner, when the face region image is input from the face detection unit 102, the face recognition unit 103 calculates a similarity score for each registered face image. Furthermore, the face recognition unit 103 identifies the maximum similarity score among the plurality of calculated similarity scores. If the maximum similarity score is greater than a predetermined threshold value (referred to as a “recognition threshold”), the face recognition unit 103 recognizes the person ID corresponding to the registered image from which the maximum similarity score was calculated. is output as the recognition result. On the other hand, when the maximum similarity score is smaller than the predetermined recognition threshold, the face recognition unit 103 determines that the person is not one of the registered persons (determines that the person is an unregistered person), indicating that the person is an unregistered person. Information is output as a recognition result.”). Regarding claim 7, Yamamoto discloses all limitations of claim 6. Yamamoto further discloses wherein in a case where the adjusted similarity is higher than a threshold, it is determined that the target objects present in the two object area images are the same as each other (para. 0020, “In this manner, when the face region image is input from the face detection unit 102, the face recognition unit 103 calculates a similarity score for each registered face image. Furthermore, the face recognition unit 103 identifies the maximum similarity score among the plurality of calculated similarity scores. If the maximum similarity score is greater than a predetermined threshold value (referred to as a “recognition threshold”), the face recognition unit 103 recognizes the person ID corresponding to the registered image from which the maximum similarity score was calculated. is output as the recognition result. On the other hand, when the maximum similarity score is smaller than the predetermined recognition threshold, the face recognition unit 103 determines that the person is not one of the registered persons (determines that the person is an unregistered person), indicating that the person is an unregistered person. Information is output as a recognition result.”), and wherein in a case where, based on the feature information extracted from the image as the processing target and the accumulated feature information, it is determined that the image as the processing target is an inappropriate image for object recognition, the similarity is adjusted to be small (para. 0036, “Further, the information processing apparatus 100 of the present embodiment collects and accumulates score correspondence data in order to specify the "inappropriate detection condition". Then, based on the collected score correspondence data, the information processing apparatus 100 specifies shooting conditions such that the similarity score does not exceed the recognition threshold even if the target person is a registered person. In other words, the information processing apparatus 100 selects an area in which the target person's face is too small in the image because it is far from the monitoring camera 200, an area in which the target person is in a position close to the monitoring camera 200, and therefore has a large downward angle. , a region whose similarity score does not exceed the threshold”). Regarding claim 8, as best understood, Yamamoto discloses all limitations of claim 6. Yamamoto further discloses wherein in a case where it is determined that an object area image is inappropriate for object recognition, a request to register the object area image for object recognition is refused (paras. 0035-0036, “As described above, in the information processing apparatus 100 of the present embodiment, by specifying the “inappropriate detection condition”, face recognition processing is skipped for face detection results that are determined to match the condition. By doing so, it becomes possible to determine that the person is an unregistered person. By skipping the face recognition process in this way, the processing amount of the face recognition process as a whole can be reduced…Further, the information processing apparatus 100 of the present embodiment collects and accumulates score correspondence data in order to specify the "inappropriate detection condition". Then, based on the collected score correspondence data, the information processing apparatus 100 specifies shooting conditions such that the similarity score does not exceed the recognition threshold even if the target person is a registered person.”, wherein the subsequent skipping of the facial recognition process, which requires a registration attempt, constitutes a refusal-to-register for enhanced computational efficiency) . Claim Rejections - 35 USC § 103 07-06 AIA 15-10-15 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. 07-20-aia AIA 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. 07-21-aia AIA Claim s 9-10 are rejected under 35 U.S.C. 103 as being unpatentable over Yamamoto in view of Hasabnis et al. (US PG Pub 20220092042, hereinafter “Hasabnis”) . Regarding claim 9, Yamamoto discloses all limitations of claim 1. Although Yamamoto discloses feature information extraction from any number of a plurality of input images and determination of whether the image is inappropriate, Yamamoto does not disclose wherein the images are a plurality of learning images, or wherein an image determined to be an inappropriate image among the plurality of learning images is not used to train a learning model. However, Hasabnis discloses a method and system of improving the performance of an artificial intelligence training dataset, teaching the remaining limitations of claim 9, namely: wherein the images are a plurality of learning images (para. 0122, “To generate a robust AI-based model, it may be desirable to use large amount of training data. Accordingly, open-source or crowd sourced data may be used as training data. However, open-sourced and/or crowd-sourced data may include low quality and/or malicious data. …In some examples, a person may provide poor quality data corresponding to bugs, errors, and/or another poor-quality metric to a repository. Using poor quality data for AI model training may result in a poor-quality AI-based model. Examples disclosed herein provide a mechanism for determining the quality of data (e.g., datapoints) in a dataset, quality of the datasets, and/or quality of the repositories. If the quality of the training dataset is below a threshold, examples disclosed herein remove low quality data from the training dataset to improve the training dataset. ”), wherein the disclosure of Yamamoto with regard to the content of the accumulated facial recognition image feature data would be applicable as it would constitute a dataset to be used as a training dataset for the Ai-based model of Hasabnis. wherein an image determined to be an inappropriate image among the plurality of learning images is not used to train a learning model (para. 0122, “If the quality of the training dataset is below a threshold, examples disclosed herein remove low quality data from the training dataset to improve the training dataset”) , wherein the threshold or general metric to be used within the decision to remove the learning images would be the similarity metric combined with the appropriacy judgement of Yamamoto. Therefore, Hasabnis discloses a method and system of improving training an artificial intelligence model. Yamamoto discloses that, in certain embodiments, the architecture for the facial feature extraction and general recognition pipeline is a deep neural network-based architecture (para. 0009). Thus, it would have been obvious for one having ordinary skill in the art prior to the effective filing date of the claimed invention to have utilized the disclosure of Hasabnis with respect to the “best practices” of training an AI-based model for improved performance within the method and system of Yamamoto as a teaching or suggestion in the prior art which would have enabled an ordinarily skilled artisan to have modified the learning of the facial recognition model of Yamamoto to exclude inappropriate images while learning. Regarding claim 10, Yamamoto in view of Hasabnis discloses all limitations of claim 9. Yamamoto further discloses wherein: the learning images are managed in association with class labels of the learning images (para. 0009, wherein the pre-registered images each have classifications), and wherein based on representative feature information representing a plurality of pieces of feature information identified based on a plurality of pieces of feature information regarding a plurality of learning images associated with a class label and the accumulated feature information, it is determined whether the representative feature information is inappropriate for object recognition (para. 0030, “Then, when there is a predetermined number of samples or more of threshold-exceeding data in the accumulated data output from the accumulating unit 502, the range determining unit 503 extracts coordinates from the threshold-exceeding data, and extracts coordinates from the threshold-exceeding data. Identify. For example, when score correspondence data such as that shown in FIG. Extract. Then, the range determination unit 503 defines coordinates (400, 200), coordinates (400, 300), coordinates (500, 300), and coordinates (500, 200) as the vertices of a circumscribing rectangle that includes these coordinates. Specifies the rectangle to Hereinafter, the area inside the rectangle will be referred to as the "recognition target range", and the area outside the rectangle will be referred to as the "non-recognition target range". In this embodiment, the "non-recognition target range" corresponds to the "inappropriate detection condition"”, wherein the method and system disclosed by Yamamoto identifies a particular range within the photo as a suitable target (the “recognition target range”), while the unsuitable range outside of the range corresponds to the “inappropriate detection condition” for object recognition. Examiner notes that this process would be easily modifiable and of little effort for the ordinarily skilled artisan to apply to a plurality of images ). Yamamoto does not explicitly disclose wherein in a case where it is determined that the representative feature information is inappropriate, the plurality of learning images is managed so that none of all the plurality of learning images associated with the class label is used to train the learning model. However, Hasabnis discloses wherein in a case where it is determined that the representative feature information is inappropriate, the plurality of learning images is managed so that none of all the plurality of learning images associated with the class label is used to train the learning model (para. 0122, “If the quality of the training dataset is below a threshold, examples disclosed herein remove low quality data from the training dataset to improve the training dataset”. Examiner notes that this process would be easily modifiable and of little effort for the ordinarily skilled artisan to apply to a training dataset comprising a plurality of learning images, especially utilizing the inappropriate- ness metric of Yamamoto as representative of the feature information ). Thus, it would have been obvious to one having ordinary skill in the art prior to the effective filing date of the claimed invention to have combined the disclosures of Yamamoto and Hasabnis according to the rationale of claim 9. Conclusion Any inquiry concerning this communication or earlier communications from the examiner should be directed to ROHAN TEJAS MUKUNDHAN whose telephone number is (571)272-2368. The examiner can normally be reached Monday - Friday 9AM - 6PM. 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, Gregory Morse can be reached at 5712723838. 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. /ROHAN TEJAS MUKUNDHAN/Examiner, Art Unit 2663 /GREGORY A MORSE/Supervisory Patent Examiner, Art Unit 2698 Application/Control Number: 18/793,309 Page 2 Art Unit: 2663 Application/Control Number: 18/793,309 Page 3 Art Unit: 2663 Application/Control Number: 18/793,309 Page 4 Art Unit: 2663 Application/Control Number: 18/793,309 Page 5 Art Unit: 2663 Application/Control Number: 18/793,309 Page 6 Art Unit: 2663 Application/Control Number: 18/793,309 Page 7 Art Unit: 2663 Application/Control Number: 18/793,309 Page 8 Art Unit: 2663 Application/Control Number: 18/793,309 Page 9 Art Unit: 2663 Application/Control Number: 18/793,309 Page 10 Art Unit: 2663 Application/Control Number: 18/793,309 Page 11 Art Unit: 2663 Application/Control Number: 18/793,309 Page 12 Art Unit: 2663 Application/Control Number: 18/793,309 Page 13 Art Unit: 2663 Application/Control Number: 18/793,309 Page 14 Art Unit: 2663 Application/Control Number: 18/793,309 Page 15 Art Unit: 2663 Application/Control Number: 18/793,309 Page 16 Art Unit: 2663 Application/Control Number: 18/793,309 Page 17 Art Unit: 2663
Prosecution Insights