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 .
Election/Restrictions
Applicant's election of Species I (claims 2, 11, and 16) in the reply filed on 04/20/2026 is acknowledged. Applicant's election is noted with traverse.
Although Applicant points to the four shared processing steps recited in generic claims 1, 9, and 10 (extracting human movements in any number of frames; computing time-series feature values; computing similarity between time-series feature values; classifying movements based on similarity) and argues that the species are merely refinements of a single core inventive concept, the species are set forth to independent and distinct technical features such that the search is diverse for each species ( Species I is directed to computing similarity between two time-series feature values having different numbers of frames by determining, for each frame of one time-series, a corresponding frame of the other time-series based on pose-feature similarity, and then computing similarity based on pose-feature similarity of the associated frame pairs. Species II is directed to to computing similarity between two time-series feature values having different numbers of frames by extracting a plurality of key frames from one time-series, determining key-relevance frames from the other time-series based on pose-feature similarity, and computing similarity using one or more of pose similarity, time-interval similarity, change-direction similarity, and/or key-relevance frame determination result, optionally with weighted combination. Species III is directed to automated extraction of human movements from a moving image using a tracking engine that tracks same-person identity, extracting movements only when a person's consecutive appearance satisfies a lower-limit threshold, and/or dividing frames into groups when consecutive appearance exceeds an upper-limit threshold. )
Thus, while the restricted species may be disclosed as alternative embodiments within the same action classification system, the species are directed to different underlying concepts and different technical approaches: a frame-to-frame pose-correspondence algorithm for variable-length sequence alignment, a key-frame abstraction and multi-metric temporal similarity architecture, and a tracking-engine-based movement segmentation pipeline with consecutive-appearance thresholding. A search directed to one species would not necessarily be expected to identify the most relevant prior art for the other species. For example, a search for frame-to-frame sequence alignment and pose-correspondence techniques for variable-length action sequences (CPC classes G06V10/764, G06T7/20) would not necessarily locate the most relevant prior art for key-frame extraction and multi-metric temporal similarity fusion in video action retrieval; likewise, a search for key-frame-based action representation and weighted temporal similarity metrics (CPC classes G06V40/20, G06V10/74) would not necessarily locate the most relevant prior art for multi-person tracking engines, tracklet-based movement segmentation, and consecutive-appearance threshold filtering (CPC classes G06V20/52, G06T7/70).
Therefore, the search for the generic claims would not reasonably encompass the specific subject matter of each dependent species, and examination of all species together would impose a serious search and examination burden. For at least these reasons, and upon reconsideration of Applicant’s traversal, the restriction requirement is still deemed proper and is therefore made FINAL.
Accordingly, examination will proceed on the elected species only. The application has pending claims 1–20 (non-elected claims 3–8, 12–15, and 17–20 are withdrawn from further consideration).
Claim Objections
Claims 2, 11, and 16 objected to because of the following informalities: Claims 2, 11, and 16 recited the phrase “associated to”, which is grammatically incorrect. Applicant should amend the phrase to “associated with”. Appropriate correction is required.
Claim 16 is objected to because of the following informalities: Claim 16 recites “--wherein the program causing the computer to--”, which is grammatically incorrect. It should likely be amended to "--wherein the program causes the computer to-- " to establish proper active phrasing. Appropriate correction is required.
Claim Rejections - 35 USC § 112(b)
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 2, 11, and 16 are 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.
Claims 2, 11, and 16 recite “the two time-series feature values”, “the one time-series feature value”, and “the other time-series feature value”. However, parent claims 1, 9, and 10 recite computing a similarity between “a plurality of the time-series feature values” and do not provide proper antecedent basis for “the two time-series feature values”, “the one time-series feature value”, or “the other time-series feature value”. As a result, it is unclear which two time-series feature values from the plurality are being referenced and compared.
Appropriate correction may include reciting “two time-series feature values selected from the plurality of time-series feature values”, “one of the two time-series feature values”, and “the other of the two time-series feature values” .etc.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-2, 9-11, and 16 are rejected under 35 U.S.C. § 101 because the claimed invention is directed to a judicial exception without significantly more.
This rejection has been made in accordance with the current USPTO subject matter eligibility framework, including MPEP §§ 2103–2106.07, the 2019 Revised Patent Subject Matter Eligibility Guidance, the October 2019 Patent Eligibility Guidance Update, the 2024 Guidance Update on Patent Subject Matter Eligibility, Including on Artificial Intelligence, the July 2024 AI Subject Matter Eligibility Examples, the August 4, 2025 USPTO memorandum titled “Reminders on evaluating subject matter eligibility of claims under 35 U.S.C. § 101”, and the USPTO’s guidance concerning Ex parte Desjardins, Appeal No. 2024-000567. The claims have been evaluated under the broadest reasonable interpretation, and the claims have been considered as a whole.
Step 1: Statutory category
Independent claim 1 is directed to an apparatus comprising a memory and processor and therefore falls within the statutory category of a machine. Independent claim 9 is directed to a method and therefore falls within the statutory category of a process. Independent claim 10 is directed to a non-transitory storage medium storing a program and therefore falls within the statutory category of a manufacture. Accordingly, the analysis proceeds to Step 2A.
Step 2A, Prong One (Judicial exception)
Independent claim 1 recites: “extract, from a moving image, a plurality of human movements presented in any number of frames”; “compute a time-series feature value for any number of frames by computing, for each of the extracted human movements, a feature value of a human pose in each of the any number of frames”; “compute a similarity between a plurality of the time-series feature values”; and “classifying a plurality of extracted human movements, based on the similarity”. These limitations recite an abstract idea, namely collecting motion information, evaluating features of that motion, performing mathematical similarity calculations, and categorizing the movements based on those calculations.
The claim recites information collection, information evaluation, mathematical correlation, and information classification. The "human movements", "time-series feature value", "feature value of a human pose", and "similarity" are used as information items in a data-classification process. The claim does not recite an improvement to the way video frames are captured, how moving images are digitized, how pixels are encoded or compressed, or how a sensor physically captures body mechanics. Rather, generic video image information is collected, mathematically evaluated for feature values, mathematically compared, and categorized.
The claim is similar in character to claims that courts have found abstract where the focus is collecting information, mathematically analyzing the information, and presenting or acting on the results of the analysis. In Electric Power Group, LLC v. Alstom S.A., the Federal Circuit recognized claims directed to collecting and analyzing information as abstract. The present claims similarly collect movement information from frames, analyze the feature values of poses mathematically, compute a similarity score, and classify the result.
Claims 2, 11, and 16 further recite, when computing a similarity between two time-series feature values for different numbers of frames, determining a frame of the other time-series feature value associated to each frame of the one time-series feature value based on a similarity of a feature value of a human pose in each frame, and computing a similarity between the two time-series feature values based on a similarity of feature values of human poses in frames associated to each other. These limitations recite additional abstract mathematical concepts and mental processes, including calculating distance/similarity between data points, mapping or associating data points based on that mathematical correlation, and generating an aggregated similarity score.
The score-generating and frame-association limitations recite mathematical concepts. Visual comparison of movements has historically been performed by humans by observing poses and judging similarity. The claims merely recite performing such comparison and categorization using computed feature values and mathematical similarity scores without reciting a particular technological improvement to the computer performing the comparison.
The character of the elected claims as a whole is video data collection, mathematical feature computation, frame-to-frame similarity scoring, and classification logic, not an improvement to video-capture technology, computer memory, network operation, image compression, or machine-learning model architecture itself. Independent claims 9 and 10 recite substantially the same abstract idea in method and computer-program-product form. Merely implementing the same abstract mathematical and classification process using generic computer components does not avoid the judicial exception.
Accordingly, claims 1-2, 9- 11, and 16 recite an abstract idea under Step 2A, Prong One.
Step 2A, Prong Two (Practical Application)
The additional elements, considered individually and in combination, do not integrate the abstract idea into a practical application.
The recited "at least one memory", "at least one processor", "moving image", and "frames" amount to data objects, generic computer components, and generic computer implementation of the abstract classification and similarity computation concept.
The claims do not recite a particular improvement to video-capture or image-processing technology. They do not improve how a frame is sensed, sampled, imaged, digitized, segmented, filtered, encoded, compressed, or transmitted. The claims merely require extracting movements from generic frames. The claims also do not recite a particular improvement to physical tracking technology. They do not recite a new sensor arrangement, a new depth-sensing sampling technique, or a new hardware-based process for extracting skeletal features. Rather, the claims use frame data as input for mathematical comparison, scoring, and classification.
The claims further do not recite a particular improvement to artificial-intelligence or machine-learning technology. The claims do not train a model, update model parameters, modify model architecture, reduce model complexity, reduce model storage, preserve prior model knowledge, improve inference speed, improve model accuracy by a specific claimed technical mechanism, or otherwise improve how an AI classifier operates. The classification is recited purely functionally as an output of the similarity computation.
This analysis is consistent with the USPTO’s 2024 AI subject matter eligibility guidance and AI examples, which emphasize that AI-related claims may be eligible when they recite a specific technological improvement or otherwise integrate a judicial exception into a practical application. The present claims do NOT recite such a specific technological improvement. Instead, the claims use generic computer operations to collect frame information, compute numerical feature values, compare those values, and output a classification result.
This case is distinguishable from Ex parte Desjardins. In Desjardins, the claims were found to reflect an improvement in machine-learning technology itself, including training a machine-learning model on a series of tasks while preserving prior knowledge and reducing complexity/storage burdens. Here, the claims do NOT recite a particular training technique, model architecture, parameter-update mechanism, memory-saving arrangement, or data structure that improves the operation of a machine-learning model. The claimed feature values and frame-to-frame similarity mapping merely determine the ultimate abstract classification.
Nor does limiting the abstract idea to the environment of action or movement classification make the claims eligible. In Recentive Analytics, Inc. v. Fox Corp., the Federal Circuit rejected the argument that applying machine learning to a new field of use was sufficient for eligibility where the claims did not recite a technical improvement to the machine-learning process itself. Similarly here, applying mathematical similarity algorithms and classification to human poses in video frames is a field-of-use limitation, not an integration of the abstract idea into a practical application.
Accordingly, the claims do not integrate the judicial exception into a practical application under Step 2A, Prong Two.
Step 2B: (Inventive Concept)
The additional elements, considered both individually and as an ordered combination, do not amount to significantly more than the abstract idea.
The claims use generic computer components to perform ordinary computer functions, including receiving image frames, computing values, associating data points, and classifying data. These are conventional data-processing operations performed using generic computer technology.
The ordered combination also does not provide an inventive concept. The ordered combination follows the abstract idea itself: extract a movement from frames, compute pose feature values, associate frames of differing lengths based on feature similarity, compute an overall similarity, and classify the movement based on the similarity. This is no more than the abstract idea implemented on generic computer components.
Dependent claims 2, 11, and 16 recite additional steps of associating individual frames of one sequence to another and computing overall similarity based on those associations. These limitations merely specify the mathematical mechanics used in the abstract evaluation (e.g., Dynamic Time Warping logic) and do not add significantly more.
Claims 9 and 10 recite method and computer program product counterparts using processors and memories to perform substantially the same operations as apparatus claim 1. The recitation of generic processors and memories does not transform the abstract idea into patent-eligible subject matter.
Accordingly, claims 1-2, 9-11, and 16 are directed to a judicial exception without significantly more and are therefore rejected under 35 U.S.C. § 101.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (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.
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.
Claims 1–2, 9–11 and 16 are rejected under 35 U.S.C. 103 as being unpatentable over Luo (Luo et al, US 2021/0271892 A1, 2021).
Regarding claim 1, Luo teaches an action classification apparatus comprising:
at least one memory configured to store one or more instructions; and
at least one processor configured to execute the one or more instructions to:
( [Fig. 11-12], [0260-0263]: Luo teaches an electronic device for action recognition, wherein the electronic device includes one or more processors and memory storing computer-readable instructions or one or more programs that, when executed by the one or more processors, cause the one or more processors to perform the action recognition operations. Luo further teaches that the electronic device is also a computer device or computing device. )
extract, from a moving image, a plurality of human movements presented in any number of frames;
( [Fig. 1-2 & 7], [0093-0114]: Luo teaches obtaining a target video and extracting a plurality of target windows from the target video, each target window including a plurality of consecutive video frames. Luo further teaches that first action feature information of each target window describes a dynamic action included in the target window, and that the first action feature information comprises movement of one or more body parts of a subject. )
compute a time-series feature value for any number of frames by computing, for each of the extracted human movements, a feature value of a human pose in each of the any number of frames;
( [Fig. 4-5 & 7], [0127-0152]: Luo teaches extracting, for each video frame in a target window, a plurality of body key points in the video frame, and performing action recognition on the video frame according to a distribution of the body key points to obtain second action feature information of the video frame. Luo teaches that the second action feature information may include angles between body parts, displacement amounts of body key points relative to a reference video frame, and ratios between body parts. Luo further teaches combining the second action feature information of the video frames in the target window to obtain first action feature information/ first action matrix for the target window.)
computes compute a similarity between a plurality of the time-series feature values; and
( [Fig. 1- Step S106 and Fig. 5-7], [0153-0177]: Luo teaches obtaining a similarity between first action feature information of each target window and preset feature information. Luo further teaches that the first action feature information may be a first action matrix including M first action vectors, and the preset feature information may be a preset action matrix including N preset action vectors. Luo creates a similarity matrix and obtains similarities between first action vectors and preset action vectors, and determines a similarity between the first action feature information and the preset feature information. Luo also teaches a similarity sequence for multiple target windows. )
classify a plurality of extracted human movements, based on the similarity.
( [Fig. 1- Step S107 and Fig. 6-7], [0178-0198]: Luo teaches determining, from the obtained similarities corresponding to the plurality of target windows, a highest first similarity and a first target window corresponding to the highest first similarity. Luo further teaches determining/ classifying the dynamic action corresponding to the highest first similarity as the preset dynamic action when threshold conditions are satisfied. This classifies the plurality of extracted human movements, the target windows, as either containing or not containing the preset dynamic action, based on the computed similarities.)
Although different embodiments of Luo have been referred to, it would have been exceedingly obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Luo by combining Luo’s similar embodiments in order to not limit the embodiments to themselves but include other evident combinations and extensions thereof (see Luo’s [multiple paragraphs]: “In some embodiments --”).
Regarding claim 2, Luo teaches the action classification apparatus according to claim 1, wherein the at least one processor is further configured to execute the one or more instructions to,
when computing a similarity between the two time-series feature values for different numbers of frames,
determine a frame of the other time-series feature value associated to each frame of the one time-series feature value, based on a similarity of a feature value of a human pose in each frame, and
compute a similarity between the two time-series feature values, based on a similarity of feature values of human poses in frames associated to each other.
( [Fig. 5], [0024-0026], [0055-0057], [0158-0177], [0193-0198]: Luo teaches that first action feature information is a first action matrix including M first action vectors, and preset feature information is a preset action matrix including N preset action vectors, where M and N are positive integers. Luo further teaches examples in which M is 30 and N is 20, thereby teaching comparison between two action/ time-series feature values having different numbers of frame/ action vectors. Luo creates a similarity matrix having M rows and N columns or N rows and M columns. Each specified position in the similarity matrix corresponds to an i-th first action vector and a j-th preset action vector, and Luo obtains a similarity for that position based on a similarity between the i-th first action vector and the j-th preset action vector. Luo’s first action vectors and preset action vectors are generated from per-frame body-key-point/ action feature values, including body-part angles, body-key-point displacement amounts, and body-part distance ratios. Luo further teaches that the similarity at a matrix position represents similarity between the dynamic action up to the i-th frame in the target video and the preset dynamic action up to the j-th frame in a video in which the preset dynamic action is performed. Luo computes the similarity of the position corresponding to the M-th first action vector and the N-th preset action vector as the similarity between the first action feature information and the preset feature information, and provides DTW-style code using insertion, deletion, and match operations.)
Regarding claims 9–11 and 16, the rationale provided in the rejection of claims 1–2 is incorporated herein. In addition, the action classification apparatus of claims 1–2 corresponds to the methods of claims 9 and 11, as well as the non-transitory storage medium of claims 10 and 16, and performs the steps disclosed herein. Therefore, the claims are all rejected.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to KEN KUDO whose telephone number is (571)272-4498. The examiner can normally be reached M-F 8am - 5pm.
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, Vincent Rudolph can be reached at 571-272-8243. 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.
KEN KUDO
Examiner
Art Unit 2671
/KEN KUDO/Examiner, Art Unit 2671
/VINCENT RUDOLPH/Supervisory Patent Examiner, Art Unit 2671
Prosecution Insights