CTNF 17/932,339 CTNF 80135 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. DETAILED ACTION This is a Non-Final Office Action , in responses to Applicant’s RCE filed 07/22/2026 . It is noted; the current Patent Application was filed 09/15/2022 . Claim(s) 1-25 are pending. Claim(s) 1, 7, 14 and 20 are independent claims. 07-50 AIA The indicated allowability of claim (s) 1-25 is withdrawn in view of the newly discovered reference(s) to Claim(s) 1-25 . Rejections based on the newly cited reference(s) follow. 07-42-05 AIA A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after allowance or after an Office action under Ex Parte Quayle , 25 USPQ 74, 453 O.G. 213 (Comm'r Pat. 1935). Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, prosecution in this application has been reopened pursuant to 37 CFR 1.114. Applicant's submission filed on 07/22/2026 has been entered. 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. Information Disclosure Statement A signed and dated copy of applicant’s IDS, which was filed 07/20/2026 is/are attached to this Office Action. Claims Rejection – 35 U.S.C. 103 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 of this title, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made . 07-21-aia AIA Claim (s) 1-25 rejected under 35 U.S.C. 103 as being unpatentable over Li et al., NPL (“MS-TCN++: Multi-Stage Temporal Convolutional Network for Action Segmentation” Published 09/02/2020 (13 pages) by IEEE [hereinafter “Li”], in view of Julio et al., (“US20210081756 A1” Published 03/18/2021 filed 11/30/2020 [hereinafter “Julio”] . Independent Claim 1, Li teaches : A computing system comprising: a data storage to store input data that includes a plurality of portions, wherein a subset of the plurality of portions collectively represents a first action; (In Li the Abstract and the Introduction pages 1-2, discloses vector time series representation of videos to model the temporal dynamics of cornplex actions using methods from linear dynamical systems theory, including a deep learning in classifying short trimmed videos, more attention has been focused on temporally segmenting and classifying activities in long untrimmed videos…With the success of temporal convolutional networks (TCNs) as a powerful temporal model for speech synthesis, many researchers adapt TCN-based models for the temporal action segmentation task [11], [12], [13]. These models were more capable in capturing long range dependencies between the video frames by relying on a large receptive field…) Li further teaches and a controller implemented in one or more of configurable logic OR fixed-functionality logic , wherein the controller is to, during training of a neural network; (In Li the Abstract , discloses deep learning in classifying short trimmed videos, more attention has been focused on temporally segmenting and classifying activities in long untrimmed videos…and further in Li section 3.5 pages 5-6 ; mentioning the configurable logic … i.e., the loss function, wherein the loss function and a cross entropy loss uses a combination of a classification loss and a smoothing loss…) Li further teaches identify, with the neural network, that a predetermined amount of the first action is completed at a first portion of the plurality of portions; and generate a first loss based on the predetermined amount of the first action being identified as being completed at the first portion; (In Li the Abstract , and the introduction pages 1-2 and section 3.1 to 3.3 pages 3-5, discloses deep learning in classifying short trimmed videos,…wherein, the task(s) is/are completed in a rnulti-stages architectures; which is composing several models sequentially such that each model operates directly on the output of the previous one; the network is able to capture dependencies between action classes and learn plausible action sequences, which helps in reducing the over-segmentation errors..) It is noted, Li’s loss function predicted probability for the ground truth label “c” at time “t”…. While the cross-entropy loss already performs well, …wherein the predictions for some of the videos contain a few over-segmentation errors. To further improve the quality of the predictions, that uses an additional smoothing loss to reduce such over segmentation errors. For this loss, uses a truncated mean squared error over the frame-wise log-probabilities…then concatenated the output of the last dilated convolutional layer in each stage to the input probabilities of the next stage (see Li section 3.1 to 3.3 pages 3-5 and section(s) 4.3-4.3 pages 7-8) However, Li does not expressly teach, But the combination of Li and Julio teach ... … and update the neural network based on the first loss, wherein updating includes updating one or more weights, activation functions, OR biases . (In Julio Para 65 , discloses once the neural network is structured, a learning model can be applied to the network to train the network to perform specific tasks. The learning model describes how to adjust the weights within the model to reduce the output error of the network. Backpropagation of errors is a common method used to train neural networks. An input vector is presented to the network for processing. The output of the network is compared to the desired output using a loss function and an error value is calculated for each of the neurons in the output layer . The error values are then propagated backwards until each neuron has an associated error value which roughly represents its contribution to the original output. The network can then learn from those errors using an algorithm, such as the stochastic gradient descent algorithm, to update the weights of the of the neural network . In addition, Julio Para 124 , further includes The processor cores 1118 and number of hardwired or configurable circuits, some or all of which may include programmable and/or configurable combinations of electronic components, semiconductor devices, and/or logic elements that are disposed partially or wholly in a PC, server, or other computing system capable of executing processor-readable instructions.) Accordingly, it would have been obvious to one having ordinary skill in the art at the time before the effective filing date of the claimed invention was made to modify Li’s loss function and deep learning in classifying short trimmed videos, more attention has been focused on temporally segmenting and classifying activities in long untrimmed videos, to include a means of said, update the neural network based on the first loss, wherein updating includes updating one or more weights, activation functions, OR biases as taught by Julio, that provides Artificial intelligence (AI), including machine learning (ML), deep learning (DL), and/or other artificial machine-driven logic, enables machines (e.g., computers, logic circuits, etc.) to use a model to process input data to generate an output based on patterns and/or associations previously learned by the model via a training process. For instance, the model may be trained with data to recognize patterns and/or associations and follow such patterns and/or associations when processing input data such that other input(s) result in output(s) consistent with the recognized patterns and/or associations (in Julio Para 18). It is noted the KSR ruling recommends references directed to similar subject matter to be combined. Claim 2, Li and Julio further teach : generate a first vector based on the subset of the plurality of portions and the predetermined amount of the first action being identified as being completed at the first portion, and identify a second vector that is a ground truth, wherein to generate the first loss, the controller is to compare the first vector to the second vector; (In Li section 3.2 page 4 , introduces a multi-stage temporal convolutional network (MSTCN) for the temporal action segmentation task. In this multi-stage model, each stage takes an initial prediction from the previous stage and refines it. The input of the first stage are the frame-wise features of the video as shown in equation(s) (5) and (6)…wherein a loss function is used a combination of a classification loss and a smoothing loss. For the classification loss; used a cross-entropy loss (11) …wherein loss function predicted probability for the ground truth label “c” at time “t” (i.e., SVM Support Vector Machine) …. While the cross-entropy loss already performs well, …wherein the predictions for some of the videos contain a few over-segmentation errors. To further improve the quality of the predictions, that uses an additional smoothing loss to reduce such over segmentation errors. For this loss, uses a truncated mean squared error over the frame-wise log-probabilities…then concatenated the output of the last dilated convolutional layer in each stage to the input probabilities of the next stage (see Li section 3.1 to 3.3 pages 3-5 and section(s) 4.3-4.3 pages 7-8). In addition, Julio in Para 65 , mentions once the neural network is structured, a learning model can be applied to the network to train the network to perform specific tasks. The learning model describes how to adjust the weights within the model to reduce the output error of the network. Backpropagation of errors is a common method used to train neural networks. An input vector is presented to the network for processing. The output of the network is compared to the desired output using a loss function and an error value is calculated for each of the neurons in the output layer . The error values are then propagated backwards until each neuron has an associated error value which roughly represents its contribution to the original output. The network can then learn from those errors using an algorithm, such as the stochastic gradient descent algorithm, to update the weights of the of the neural network. Accordingly, it would have been obvious to one having ordinary skill in the art at the time before the effective filing date of the claimed invention was made to modify Li’s loss function and deep learning in classifying short trimmed videos, more attention has been focused on temporally segmenting and classifying activities in long untrimmed videos, to include a means of said, generate a first vector based on the subset of the plurality of portions and the predetermined amount of the first action being identified as being completed at the first portion,… as taught by Julio, that provides Artificial intelligence (AI), including machine learning (ML), deep learning (DL), and/or other artificial machine-driven logic, enables machines (e.g., computers, logic circuits, etc.) to use a model to process input data to generate an output based on patterns and/or associations previously learned by the model via a training process. For instance, the model may be trained with data to recognize patterns and/or associations and follow such patterns and/or associations when processing input data such that other input(s) result in output(s) consistent with the recognized patterns and/or associations (in Julio Para 18). It is noted the KSR ruling recommends references directed to similar subject matter to be combined. Claim 3, Li and Julio further teach : process, with the neural network, the subset of the plurality of portions to generate an output, wherein to generate the first vector, the controller is to execute, with a plurality of convolution layers, a plurality of convolutions on the output; (In Li section 3.2 page 4 , introduces a multi-stage temporal convolutional network (MSTCN) for the temporal action segmentation task. In this multi-stage model, each stage takes an initial prediction from the previous stage and refines it. The input of the first stage are the frame-wise features of the video as shown in equation(s) (5) and (6)…wherein a loss function is used a combination of a classification loss and a smoothing loss. For the classification loss; used a cross- entropy loss (11) …wherein loss function predicted probability for the ground truth label “c” at time “t” (i.e., SVM Support Vector Machine) …. While the cross-entropy loss already performs well, …wherein the predictions for some of the videos contain a few over-segmentation errors. To further improve the quality of the predictions, that uses an additional smoothing loss to reduce such over segmentation errors. For this loss, uses a truncated mean squared error over the frame-wise log-probabilities…then concatenated the output of the last dilated convolutional layer in each stage to the input probabilities of the next stage (see Li section 3.1 to 3.3 pages 3-5 and section(s) 4.3-4.3 pages 7-8). Claim 4, Li and Julio further teach : wherein a plurality of residual connections connects the plurality of convolution layers; (In Li section 3.2 page 4 , introduces a multi-stage temporal convolutional network (MSTCN) for the temporal action segmentation task. In this multi-stage model, each stage takes an initial prediction from the previous stage and refines it. The input of the first stage are the frame-wise features of the video as shown in equation(s) (5) and (6).) Claim 5, Li and Julio further teach : process, with the neural network, the plurality of portions to identify segments that correspond to a plurality of actions and label the segments with action labels, wherein the plurality of actions includes the first action, generate a second loss based on the segments and the action labels, and update the neural network based on the second loss; (In Li section 3.2 page 4 , introduces a multi-stage temporal convolutional network (MSTCN) for the temporal action segmentation task. In this multi-stage model, each stage takes an initial prediction from the previous stage and refines it. The input of the first stage are the frame-wise features of the video as shown in equation(s) (5) and (6). Moreover, ( Li section 3.1 to 3.3 pages 3-5 and section(s) 4.3-4.4 pages 7-8, further mentions the rnulti-stage architectures; is composing several models sequentially such that each model operates directly on the output of the previous one… since the output of each stage is an initial prediction, the network is able to capture dependencies between action classes and learn plausible action sequences. Also, in page 2 first paragraph and page 3 first paragraph ., further mentions the identify segments that correspond to a plurality of actions and label the segments with action labels.) In addition, Julio in Para 78 , mentions Variations on supervised and unsupervised training may also be employed. Semi-supervised learning is a technique in which in the training dataset 502 includes a mix of labeled and unlabeled data of the same distribution . Incremental learning is a variant of supervised learning in which input data is continuously used to further train the model. Incremental learning enables the trained neural network 508 to adapt to the new data 512 without forgetting the knowledge instilled within the network during initial training….) Accordingly, it would have been obvious to one having ordinary skill in the art at the time before the effective filing date of the claimed invention was made to modify Li’s loss function and deep learning in classifying short trimmed videos, more attention has been focused on temporally segmenting and classifying activities in long untrimmed videos, to include a means of said, label the segments with action labels, wherein the plurality of actions includes the first action, generate a second loss based on the segments and the action labels,… as taught by Julio, that provides Artificial intelligence (AI), including machine learning (ML), deep learning (DL), and/or other artificial machine-driven logic, enables machines (e.g., computers, logic circuits, etc.) to use a model to process input data to generate an output based on patterns and/or associations previously learned by the model via a training process. For instance, the model may be trained with data to recognize patterns and/or associations and follow such patterns and/or associations when processing input data such that other input(s) result in output(s) consistent with the recognized patterns and/or associations (in Julio Para 18). It is noted the KSR ruling recommends references directed to similar subject matter to be combined. Claim 6, Li and Chu further teach : wherein: the controller is to identify, with a convolutional neural network, features of the plurality of portions; the input data is one or more of video data or audio data; the neural network is a temporal convolutional network; and to identify that the predetermined amount of the first action is completed at the first portion, the controller is to process the features with the temporal convolutional network; In Li the Abstract , and the introduction pages 1-2 and section 3.1 to 3.3 pages 3-5, discloses deep learning in classifying short trimmed videos,…wherein, the task(s) is/are completed in a rnulti-stages architectures; which is composing several models sequentially such that each model operates directly on the output of the previous one; the network is able to capture dependencies between action classes and learn plausible action sequences, which helps in reducing the over-segmentation errors.. It is noted, Li’s loss function predicted probability for the ground truth label “c” at time “t”…. While the cross-entropy loss already performs well, …wherein the predictions for some of the videos contain a few over-segmentation errors. To further improve the quality of the predictions, that uses an additional smoothing loss to reduce such over segmentation errors. For this loss, uses a truncated mean squared error over the frame-wise log-probabilities…then concatenated the output of the last dilated convolutional layer in each stage to the input probabilities of the next stage (see Li section 3.1 to 3.3 pages 3-5 and section(s) 4.3-4.4 pages 7-8). wherein the predetermined amount corresponds to a midpoint of the first action. ( Li section 3.1 to 3.3 pages 3-5 and section(s) 4.3-4.4 pages 7-8, further mentions the rnulti-stage architectures; is composing several models sequentially such that each model operates directly on the output of the previous one… since the output of each stage is an initial prediction, the network is able to capture dependencies between action classes and learn plausible action sequences..(Li section 4.3-4.4 Comparing Different Loss Functions pages 7-8, mentions the loss function, uses a combination or a cross-entropy loss, which is common practice for classification tasks, and a truncated mean squared loss over the frame-wise log-probabilities to ensure smooth predictions… drop in performance is due to the fact that the smoothing l oss heavily penalizes changes in frame-wise labels, which affects the detected boundaries between action segments (in the BRI, is recognized as predetermined amount corresponds to a midpoint of the first action as claimed.) Regarding Claim(s) 7-12 (respectively) is/are fully incorporated similar subject of claim(s) 1-6 (respectively) cited above and further in view of the following: A semiconductor apparatus, the semiconductor apparatus comprising: one or more substrates; and logic coupled to the one or more substrates, wherein the logic is implemented in one or more of configurable logic OR fixed-functionality logic , the logic coupled to the one or more substrates to, during training of a neural network …(in Julio Para 124 , i.e., The processor cores 1118 may include any number of hardwired or configurable circuits, some or all of which may include programmable and/or configurable combinations of electronic components, semiconductor devices, and/or logic elements that are disposed partially or wholly in a PC, server, or other computing system capable of executing processor-readable instructions and Julio Para 65 , i.e., mentions once the neural network is structured, a learning model can be applied to the network to train the network to perform specific tasks. …to update the weights of the of the neural network.) Accordingly, it would have been obvious to one having ordinary skill in the art at the time before the effective filing date of the claimed invention was made to modify Li’s loss function and deep learning in classifying short trimmed videos, more attention has been focused on temporally segmenting and classifying activities in long untrimmed videos, to include a means of said, semiconductor apparatus, the semiconductor apparatus comprising: one or more substrates; and logic coupled to the one or more substrates, wherein the logic is implemented in one or more of configurable logic OR fixed-functionality logic, the logic coupled to the one or more substrates to, during training of a neural network,… as taught by Julio, that provides Artificial intelligence (AI), including machine learning (ML), deep learning (DL), and/or other artificial machine-driven logic, enables machines (e.g., computers, logic circuits, etc.) to use a model to process input data to generate an output based on patterns and/or associations previously learned by the model via a training process. For instance, the model may be trained with data to recognize patterns and/or associations and follow such patterns and/or associations when processing input data such that other input(s) result in output(s) consistent with the recognized patterns and/or associations (in Julio Para 18). It is noted the KSR ruling recommends references directed to similar subject matter to be combined. Regarding Claim 13 is/are fully incorporated similar subject of claim 1 cited above, and further in view of the following: wherein the logic coupled to the one or more substrates includes transistor channel regions that are positioned within the one or more substrates …(in Julio Para 124 , i.e., The processor cores 1118 may include any number of hardwired or configurable circuits, some or all of which may include programmable and/or configurable combinations of electronic components, semiconductor devices, and/or logic elements that are disposed partially or wholly in a PC, server, or other computing system capable of executing processor-readable instructions and Julio Para 65 , i.e., mentions once the neural network is structured, a learning model can be applied to the network to train the network to perform specific tasks. …to update the weights of the of the neural network.) Accordingly, it would have been obvious to one having ordinary skill in the art at the time before the effective filing date of the claimed invention was made to modify Li’s loss function and deep learning in classifying short trimmed videos, more attention has been focused on temporally segmenting and classifying activities in long untrimmed videos, to include a means of said, the logic coupled to the one or more substrates includes transistor channel regions that are positioned within the one or more substrates,… as taught by Julio, that provides Artificial intelligence (AI), including machine learning (ML), deep learning (DL), and/or other artificial machine-driven logic, enables machines (e.g., computers, logic circuits, etc.) to use a model to process input data to generate an output based on patterns and/or associations previously learned by the model via a training process. For instance, the model may be trained with data to recognize patterns and/or associations and follow such patterns and/or associations when processing input data such that other input(s) result in output(s) consistent with the recognized patterns and/or associations (in Julio Para 18). It is noted the KSR ruling recommends references directed to similar subject matter to be combined. Regarding Claim(s) 14-19 (respectively) is/are fully incorporated similar subject of claim(s) 1-6 (respectively) cited above. Regarding Claim(s) 20-25 (respectively) is/are fully incorporated similar subject of claim(s) 1-6 (respectively) cited above. Conclusion Any inquiry concerning this communication or earlier communications from the examiner should be directed to QUOC A TRAN whose telephone number is (571)272-8664. The examiner can normally be reached Monday-Friday 9am-5pm EST. 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, Cesar Paula can be reached at 571-272-4128. The fax phone number for the organization where this application or proceeding is assigned is 5`71-273-8300. 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. /QUOC A TRAN/Primary Examiner, Art Unit 2145 Application/Control Number: 17/932,339 Page 2 Art Unit: 2145 Application/Control Number: 17/932,339 Page 3 Art Unit: 2145 Application/Control Number: 17/932,339 Page 4 Art Unit: 2145 Application/Control Number: 17/932,339 Page 5 Art Unit: 2145 Application/Control Number: 17/932,339 Page 6 Art Unit: 2145 Application/Control Number: 17/932,339 Page 7 Art Unit: 2145 Application/Control Number: 17/932,339 Page 8 Art Unit: 2145 Application/Control Number: 17/932,339 Page 9 Art Unit: 2145 Application/Control Number: 17/932,339 Page 10 Art Unit: 2145 Application/Control Number: 17/932,339 Page 11 Art Unit: 2145 Application/Control Number: 17/932,339 Page 12 Art Unit: 2145 Application/Control Number: 17/932,339 Page 13 Art Unit: 2145 Application/Control Number: 17/932,339 Page 14 Art Unit: 2145 Application/Control Number: 17/932,339 Page 15 Art Unit: 2145 Application/Control Number: 17/932,339 Page 16 Art Unit: 2145
Prosecution Insights