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 .
Status of Claims
The present application is being examined under the claims filed 04/07/2025.
Claims 1-20 are pending.
Response to Amendment
This Office Action is in response to Applicant’s communication filed April 7 2025 in response to office action mailed December 12 2024 . The Applicant’s remarks and any amendments to the claims or specification have been considered with the results that follow.
Response to Arguments
Regarding 35 U.S.C. 101
In Remarks page 13, Argument 1
Former claims 1 to 20 were rejected o the assertion that they were directed to nonstatutory subject matter.
Claim 1 has been amended to recite that the output tensor is processed to perform a classification task for the batch of video data samples represented by the output tensor. By using the output tensor generated via the steps of former claim 1, the output tensor is applied to a practical problem; that is, the performance of a classification task on the batch of video data samples. It is respectfully submitted that the claim recites additional elements that amount to significantly more than the judicial exception under Step 2B of the Alice-Mayo framework.
In response to Argument 1
The examiner believes Applicant is arguing that the following limitation of amended claim 1 amounts to significantly more under step 2B:
and processing the output tensor to perform a classification task for the batch of video data samples.
MPEP 2106.04(a)(2) III. recites:
Accordingly, the "mental processes" abstract idea grouping is defined as concepts performed in the human mind, and examples of mental processes include observations, evaluations, judgments, and opinions.
The limitation amounts to performing an evaluation on an input tensor and classifying data which could, for example, be performed by a series of matrix multiplications. This could be performed in the human mind or by a human using pen and paper. Therefore, the limitation alleged to amount to significantly more
In Remarks page 13, Argument 2
Further, as noted in at least paragraph [0090] of the published application, the claimed approach makes the processing of the batch of video data samples higher in speed and accuracy, and more memory efficient. For at least these reasons, the Applicants respectfully submit that amended independent claim 1, as well as claims 2 to 9 that depend therefrom, define patentable subject matter.
In response to Argument 2,
MPEP 2106.05(a) recites:
After the examiner has consulted the specification and determined that the disclosed invention improves technology, the claim must be evaluated to ensure the claim itself reflects the disclosed improvement in technology. Intellectual Ventures I LLC v. Symantec Corp., 838 F.3d 1307, 1316, 120 USPQ2d 1353, 1359 (Fed. Cir. 2016) (patent owner argued that the claimed email filtering system improved technology by shrinking the protection gap and mooting the volume problem, but the court disagreed because the claims themselves did not have any limitations that addressed these issues). That is, the claim must include the components or steps of the invention that provide the improvement described in the specification.
[…]
It is important to note, the judicial exception alone cannot provide the improvement. The improvement can be provided by one or more additional elements. See the discussion of Diamond v. Diehr, 450 U.S. 175, 187 and 191-92, 209 USPQ 1, 10 (1981)) in subsection II, below. In addition, the improvement can be provided by the additional element(s) in combination with the recited judicial exception.
Applicant has not shown how any alleged technical improvement is reflected in the claims. Additionally, applicant has not shown how the improvement is provided by the additional elements or the additional elements in combination with the judicial exception. Therefore, the rejections of claim 1 and its dependents are maintained and updated in view of amendments to the claims. See rejections under 35 U.S.C. 101 below for a complete analysis.
In Remarks page 13, Argument 3
Independent claim 10 has been amended in a similar manner as has independent claim 1. Accordingly, the Applicants respectfully submit that amended independent claim 10, as well as 13 claims 11 to 18 that depend therefrom, also define patentable subject matter.
Independent claims 19 and 20 include, by reference, all of the limitations of amended independent claim 1. Accordingly, the Applicants respectfully submit that claims 19 and 20 also define patentable subject matter.
Withdrawal of the rejections under 35 USC § 101 is respectfully requested.
In response to Argument 3,
For the reasons given above and in the rejections under 35 U.S.C. 101 that follow, the rejections of claims 1, 10, and all dependent claims are maintained and updated in view of the amendments to the claims.
Regarding 35 U.S.C. 112
In Remarks page 14, Argument 4
Former claim 11 was rejected on the basis that this former claim was indefinite. Claim 11 has been amended to recite "generating the 4D temporal input tensor ... to form the 4D temporal input tensor''. The Applicants respectfully submit that the claim as amended is now clear and definite. Withdrawal of the rejections under 35 USC§ 112 is respectfully requested.
In response to Argument 4
Examiner agrees that the claim as amended is clear and definite under 35 U.S.C. 112(b). However, the claim amendment raises new issues under 35 U.S.C. 112(d). Notably, claim 10 specifies “a 4D temporal input tensor, based on the input tensor, having a combined spatial dimension based on the first spatial dimension and the second spatial dimension of the input tensor”. However, claim 11 specifies that “generating the 4D temporal input tensor comprises concatenating a plurality of three-dimensional (3D) tensors defined by the first spatial dimension, second spatial dimension, and input channel dimension of the input tenor to form the 4D temporal input tensor. However, concatenating these tensors would not have a combined spatial dimension as was required by claim 1. The examiner interprets the claim as though references of “4D temporal input tensor” were changed to “4D spatial input tensor”, which examiner believes is what is meant by the claim. Amending the claim this way would resolve all 35 U.S.C. 112(d) issues.
Regarding 35 U.S.C. 103
In Remarks page 15, Argument 5
The Applicants respectfully submit that the rejections are inappropriate in view of the amended claims for the reasons set forth below. It was asserted in the Office Action that Liu discloses the following feature of claim 1: obtaining the input tensor, the input tensor comprising a five-dimensional (50) tensor having ... an input channel dimension. In support of this assertion, the following passage from page 7140, column 1, paragraph 2, line 9 was cited: "Compared with it, our approach totally avoids calculating the optical flow and only requires RGC frames to train the network." The Applicants respectfully disagree with the assertion. If only a single channel is used by Liu, then there is no need for, and no mention of, a dimension corresponding to the input channel dimension, as is recited in claim 1. Thus, Liu does not teach or suggest obtaining the input tensor, the input tensor comprising a five-dimensional (5D) tensor having ... an input channel dimension.
In response to Argument 5
Examiner disagrees. Liu does teach the input tensor comprising a 5D tensor having an input channel dimension. In particular, the input channel dimension represents the color of each pixel. Each of the 5 dimensions are explained in detail below.
Dimension 1: Batch index dimension
Liu teaches dividing a video into S parts or batches {P1, P2, P3, …, Ps} then clipping each Pi to form a clipped tensor {C1, C2, …, Cs}. The batch indices are the {1,2,…s}
(page 7140 column 2 section 3.1 paragraph 2) “Formally, given a video V , we uniformly divide it into S parts[*Examiner notes: mapped to batch index dimension] {P1, P2, P3, ..., Ps }[*Examiner notes: mapped to batch of video samples] in temporal dimension. Then, a sequence of frames are chosen from Pi to form the a clip Ci.”
Dimension 2: temporal dimension
Each clip Ci contains a sequence of frames which form the temporal dimension
(page 7140 column 2 section 3.1 paragraph 2) “Formally, given a video V , we uniformly divide it into S parts {P1, P2, P3, ..., Ps } in temporal dimension. Then, a sequence of frames are chosen[*Examiner notes: mapped to temporal dimension] from Pi to form the a clip Ci.”
Dimensions 3 and 4: first and second spatial dimensions
Each frame in the clips are pictures with a width and height
(page 7142 column 2 section 4.1 line 6) “The majority of video clips in UCF101 have the 320 × 240 pixels spatial resolution[*Examiner notes: mapped to first and second spatial dimensions]”
Dimension 5: Channel dimension
Each frame of each video clip Ci is colored in red, green, and blue (RGB). Thus the input channel dimension is length 3
(page 7140 column 1 paragraph 2 line 9) “Compared with it, our approach totally avoids calculating the optical flow and only requires RGB[*Examiner notes: mapped to input channel dimension] frames to train the network.”; [*Examiner notes: Input channel dimension for videos is most often used to represent color data, i.e. RGB (red, green, and blue) pixels]
In Remarks page 15, Argument 6
Further, the Applicants respectfully submit that Liu does not teach or suggest generating a four-dimensional (4D) spatial input tensor based on the input tensor by combining the batch index dimension and temporal dimension of the input tensor into a combined batch index-temporal dimension having a size B*T [the respective sizes of the batch index dimension and the temporal dimension]. That is, the batch index dimension and the temporal dimension are collapsed into a single dimension that preserves all of the data of the two dimensions. This is explained in at least paragraph [0067] of the published application. The Applicants note that this is not as a result of processing, but simply represents a reduction in the number of dimensions without the destruction of any data.
The passage of Liu cited in the Office Action (page 7140, last paragraph) discloses the feeding forward a 3D-CNN with each clip (that is, across the temporal dimension) of the S clips, yielding S feature maps. Further, as noted in the Office Action, this passage goes on to state, "Combining S clips features with aggregating functions gains the video-level features." Liu, however simply does not teach or suggest combining the batch index dimension and temporal dimension of the input tensor into a combined batch index-temporal dimension having a size B*T. Liu processes each clip separately via its feeding forward through a 3D-CNN. Processing the data across what appears to be most similar to claim 1 's batch index dimension, by processing each clip, is quite different than collapsing the data for each clip across time.
None of the other cited prior art references rectifies these deficiencies in Liu.
As none of the cited prior art, either alone or in combination, teach or suggest all of the limitations of amended independent claim 1, the Applicants respectfully submit that amended independent claim 1, as well as claims 2 to 9 that depend therefrom, distinguish patentably over the cited prior art and should therefore be allowed.
Independent claim 10 has been amended in a similar manner as has claim 1. Accordingly, the Applicants respectfully submit that amended independent claim 10, as well as claims 11 to 18 that depend therefrom, also distinguish patentably over the cited prior art and should therefore be allowed.
Independent claims 19 and 20 include by reference all of the limitations of amended independent claim 1. Accordingly, the Applicants respectfully submit that amended independent claims 19 and 20 also distinguish patentably over the cited prior art and should therefore be allowed.
Withdrawal of the rejections under 35 USC § 103 is respectfully requested.
In response to Argument 6
Applicant’s arguments are convincing. Liu does not appear to teach the newly added features requiring the spatial input tensor to have size “B*T”, and the other cited references do not appear to cure this deficiency. However, further search revealed a new reference used in combination with the other references to teach this limitation. New grounds of rejection was necessitated by amendment.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(d):
(d) REFERENCE IN DEPENDENT FORMS.—Subject to subsection (e), a claim in dependent form shall contain a reference to a claim previously set forth and then specify a further limitation of the subject matter claimed. A claim in dependent form shall be construed to incorporate by reference all the limitations of the claim to which it refers.
The following is a quotation of pre-AIA 35 U.S.C. 112, fourth paragraph:
Subject to the following paragraph [i.e., the fifth paragraph of pre-AIA 35 U.S.C. 112], a claim in dependent form shall contain a reference to a claim previously set forth and then specify a further limitation of the subject matter claimed. A claim in dependent form shall be construed to incorporate by reference all the limitations of the claim to which it refers.
Claim 11 is rejected under 35 U.S.C. 112(d) or pre-AIA 35 U.S.C. 112, 4th paragraph, as being of improper dependent form for failing to further limit the subject matter of the claim upon which it depends, or for failing to include all the limitations of the claim upon which it depends. Applicant may cancel the claim(s), amend the claim(s) to place the claim(s) in proper dependent form, rewrite the claim(s) in independent form, or present a sufficient showing that the dependent claim(s) complies with the statutory requirements.
Regarding Claim 11
Claim 11 recites the limitation “generating the 4D temporal input tensor comprises concatenating a plurality of three-dimensional (3D) tensors defined by the first spatial dimension, second spatial dimension, and input channel dimension of the input tensor to form the 4D temporal input tensor”. However, claim 10 (the claim from which claim 11 depends) recites the limitation “generate a 4D temporal input tensor, based on the input tensor, having a combined spatial dimension based on the first spatial dimension and second spatial dimension of the input tensor”. Therefore, the temporal input tensor recited in claim 11 appears to include separate spatial dimensions and does not appear to include a combined spatial dimension based on the first spatial dimension and second spatial dimension. Therefore, the claim does not incorporate the limitations of the claim to which it refers.
For purposes of examination, the examiner will interpret the limitation of claim as though it said “generating the 4D spatial input tensor comprises concatenating a plurality of three-dimensional (3D) tensors defined by the first spatial dimension, second spatial dimension, and input channel dimension of the input tensor to form the 4D spatial input tensor”. The examiner suggests amending the claim with this language.
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 1-20 are rejected under 35 U.S.C. 101 for containing an abstract idea without significantly more.
Regarding Claim 1:
Step 1 – Is the claim to a process, machine, manufacture, or composition of matter?
Yes, the claim is to a process.
Step 2A – Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
Yes, the claim recites the abstract ideas of:
A method for processing an input tensor to generate an output tensor, comprising: — This limitation is directed to the abstract idea of a mental process (including an observation, evaluation, judgement, opinion) which can be performed by the human mind, or by a human using pen and paper (see MPEP 2106.04(a)(2) III. C.).
generating a four-dimensional ( 4D) spatial input tensor based on the input tensor by combining the batch index dimension and temporal dimension of the input tensor into a combined batch index-temporal dimension having a size B*T — This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). The claim limitations can be performed via tensor concatenation which is one kind of mathematical calculation.
performing two-dimensional (2D) convolution on the 4D spatial input tensor to generate a 4D spatial feature tensor — This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). The claim describes the mathematical operations of performing convolutions on tensors.
generating a 4D temporal input tensor, based on the input tensor, having a combined spatial dimension based on the first spatial dimension and second spatial dimension of the input tensor — This limitation is directed to the abstract idea of a mental process (including an observation, evaluation, judgement, opinion) which can be performed by the human mind, or by a human using pen and paper (see MPEP 2106.04(a)(2) III. C.).
performing 2D convolution on the 4D temporal input tensor to generate a 4D temporal feature tensor — This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). The claim describes the mathematical operations of performing convolutions on tensors.
processing the 4D spatial feature tensor and the 4D temporal feature tensor to generate the output tensor — This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). According to figure 3 box 340, the output tensor can be generated by “addition and postprocessing”. Moreover, paragraph [0032] of the specification recites “ In some embodiments, the 4D temporal feature tensor 338 and the temporally down-sampled 4D spatial feature tensor 334 are combined through elementwise addition, resulting in an output tensor 129 of the same dimensions as the 4D temporal feature tensor 338 and the temporally down-sampled 4D spatial feature tensor 334”. Thus the claim limitation is directed to performing tensor calculations using elementwise addition.
and processing the output tensor to perform a classification task for the batch of video data samples — This limitation is directed to the abstract idea of a mental process (including an observation, evaluation, judgement, opinion) which can be performed by the human mind, or by a human using pen and paper (see MPEP 2106.04(a)(2) III. C.). The limitation is directed to a mental process because it amounts to evaluating visual data (e.g. looking at an image and deciding if it contains a cat, dog or neither).
Step 2A – Prong 2 – Does the claim recite additional elements that integrate the judicial exception into a practical application?
No, the claim does not recite additional elements that integrate the judicial exception into a practical application. The additional elements:
obtaining the input tensor, the input tensor comprising a five-dimensional (5D) tensor having: a batch index dimension of size B indicating individual video data samples in a batch of video data samples; a temporal dimension of size T; a first spatial dimension; a second spatial dimension; and an input channel dimension; — This limitation is directed to mere data gathering and outputting which has been recognized by the courts (as per Ultramercial, 772 F.3d at 715, 112 USPQ2d at 1754) as insignificant extra-solution activity (see MPEP 2106.05(g)).
Step 2B – Does the claim recite additional elements that amount to significantly more than the abstract idea itself?
No, the claim does not recite additional elements which amount to significantly more than the abstract idea itself. The additional elements as identified in step 2A prong 2:
obtaining the input tensor, the input tensor comprising a five-dimensional (5D) tensor having: a batch index of size B dimension indicating individual video data samples in a batch of video data samples; a temporal dimension of size T; a first spatial dimension; a second spatial dimension; and an input channel dimension; — This limitation is recited at a high level of generality and amounts to mere data gathering of storing and retrieving information in memory, which is well-understood, routine, and conventional activity (see MPEP 2106.05(d) II.), which cannot amount to significantly more than the judicial exception.
Regarding Claim 2
Claim 2 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim is dependent on claim 1 which included an abstract idea (see rejection for claim 1). The claim recites the additional limitations:
Step 2A Prong 1:
combining the batch index dimension and temporal dimension of the input tensor comprises concatenating a plurality of three-dimensional (3D) tensors defined by the first spatial dimension, second spatial dimension, and input channel dimension to form the four-dimensional spatial input tensor — This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). The claim describes the mathematical operations of tensor concatenation in words.
Step 2A Prong 2:
and the combined batch index-temporal dimension has a size equal to the size of the batch index dimension of the input tensor multiplied by the size of the temporal dimension of the input tensor — This limitation is directed to merely limiting a judicial exception to a particular field of use (see MPEP 2106.05(h)) as it merely limits the field of the combined batch index-temporal dimension.
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2.
Step 2B:
The additional elements as identified in step 2A prong 2:
and the combined batch index-temporal dimension has a size equal to the size of the batch index dimension of the input tensor multiplied by the size of the temporal dimension of the input tensor — Merely limiting a judicial exception to a particular field of use (see MPEP 2106.05(h)) cannot amount to significantly more than the judicial exception.
Thus, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B.
Regarding Claim 3
Claim 3 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim is dependent on claim 1 which included an abstract idea (see rejection for claim 1). The claim recites the additional limitations:
Step 2A Prong 1:
generating the 4D temporal input tensor comprises concatenating a plurality of 3D tensors defined by the batch index dimension, temporal dimension, and input channel dimension of the input tensor to form the four-dimensional temporal input tensor — This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). The claim describes the mathematical operations of tensor concatenation in words.
Step 2A Prong 2:
and the combined spatial dimension has a size equal to the size of the first spatial dimension of the input tensor multiplied by the size of the second spatial dimension of the input tensor — This limitation is directed to merely limiting a judicial exception to a particular field of use (see MPEP 2106.05(h)) as it merely limits the field of the combined spatial dimension.
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2.
Step 2B:
The additional elements as identified in step 2A prong 2:
and the combined spatial dimension has a size equal to the size of the first spatial dimension of the input tensor multiplied by the size of the second spatial dimension of the input tensor — Merely limiting a judicial exception to a particular field of use (see MPEP 2106.05(h)) cannot amount to significantly more than the judicial exception.
Thus, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B.
Regarding Claim 4
Claim 4 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim is dependent on claim 1 which included an abstract idea (see rejection for claim 1). The claim recites the additional limitations:
Step 2A Prong 1:
wherein performing 2D convolution on the 4D spatial input tensor to generate the 4D spatial feature tensor comprises: applying a first number of 2D convolution filters to the 4D spatial input tensor, at a stride equal to a second number with respect to the first spatial dimension and the second spatial dimension, to generate the 4D spatial feature tensor having: a combined batch index-temporal output dimension; a first spatial output dimension, having a size equal to the first spatial dimension of the input tensor divided by the second number; a second spatial output dimension, having a size equal to the second spatial dimension of the input tensor divided by the second number; and an output channel dimension indicating a number of spatial feature maps equal to the first number. — This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). The claim describes the mathematical operations of performing convolutions on tensors.
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B.
Regarding Claim 5
Claim 5 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim is dependent on claim 4 which included an abstract idea (see rejection for claim 4). The claim recites the additional limitations:
Step 2A Prong 1:
wherein performing 2D convolution on the 4D temporal input tensor to generate the 4D temporal feature tensor comprises: applying a third number of 2D convolution filters to the 4D temporal input tensor, at a stride equal to a fourth number with respect to the temporal dimension, and a stride equal to the square of the second number with respect to the combined spatial dimension, to generate the 4D temporal feature tensor having: a batch index output dimension; a temporal dimension, having a size equal to the temporal dimension of the input tensor divided by the fourth number; a combined spatial output dimension, having a size equal to the first spatial dimension of the input tensor, divided by the second number, multiplied by the second spatial dimension of the input tensor, divided by the second number; and an output channel dimension indicating a number of temporal feature maps equal to the third number — This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). The claim describes the mathematical operations of performing tensor convolutions.
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B.
Regarding Claim 6
Claim 6 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim is dependent on claim 4 which included an abstract idea (see rejection for claim 4). The claim recites the additional limitations:
Step 2A Prong 1:
wherein processing the 4D spatial feature tensor and the 4D temporal feature tensor to generate the output tensor comprises: reshaping the 4D spatial feature tensor to generate a reshaped 4D spatial feature tensor, having: a batch index dimension having a size equal to the size of the batch index dimension of the input tensor; a temporal dimension having a size equal to the size of the temporal dimension of the input tensor; a combined spatial output dimension, having a size equal to: the first spatial dimension of the input tensor, divided by the second number, multiplied by the second spatial dimension of the input tensor, divided by the second number; and an output channel dimension indicating a number of spatial feature maps equal to the first number; — This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). The claim describes the mathematical operations of rearranging numbers into a different format and performing multiplications and divisions.
and processing the reshaped 4D spatial feature tensor and the 4D temporal 10 feature tensor to generate the output tensor — This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). According to figure 3 box 340, the output tensor can be generated by “addition and postprocessing”. Moreover, paragraph [0032] of the specification recites “In some embodiments, the 4D temporal feature tensor 338 and the temporally down-sampled 4D spatial feature tensor 334 are combined through elementwise addition, resulting in an output tensor 129 of the same dimensions as the 4D temporal feature tensor 338 and the temporally down-sampled 4D spatial feature tensor 334”. Thus the claim limitation is directed to performing tensor calculations using elementwise addition.
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B.
Regarding Claim 7
Claim 7 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim is dependent on claim 6 which included an abstract idea (see rejection for claim 6). The claim recites the additional limitations:
Step 2A Prong 1:
wherein processing the reshaped 4D spatial feature tensor and the 4D temporal feature tensor to generate the output tensor comprises: applying a number of single-element convolution filters equal to the first number to the reshaped 4D spatial feature tensor, at a temporal dimension stride equal to the fourth number with respect to the temporal dimension, to generate a temporally down-sampled 4D spatial feature tensor — This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). The claim describes the mathematical operations of performing tensor convolutions.
and processing the temporally down-sampled 4D spatial feature tensor and the 4D temporal feature tensor to generate the output tensor — This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). According to figure 3 box 340, the output tensor can be generated by “addition and postprocessing”. Moreover, paragraph [0032] of the specification recites “In some embodiments, the 4D temporal feature tensor 338 and the temporally down-sampled 4D spatial feature tensor 334 are combined through elementwise addition, resulting in an output tensor 129 of the same dimensions as the 4D temporal feature tensor 338 and the temporally down-sampled 4D spatial feature tensor 334”. Thus the claim limitation is directed to performing tensor calculations using elementwise addition.
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B.
Regarding Claim 8
Claim 8 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim is dependent on claim 7 which included an abstract idea (see rejection for claim 7). The claim recites the additional limitations:
Step 2A Prong 1:
wherein processing the temporally down-sampled 4D spatial feature tensor and the 4D temporal feature tensor to generate the output tensor comprises adding the elements of the temporally down-sampled 4D spatial feature tensor to the respective elements of the 4D temporal feature tensor to generate the output tensor —This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). The claim describes the mathematical operations of tensor addition in words.
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2. Thus, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B.
Regarding Claim 9
Claim 9 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim is dependent on claim 1 which included an abstract idea (see rejection for claim 1). The claim recites the additional limitations:
Step 2A Prong 1:
wherein: combining the batch index dimension and temporal dimension of the input tensor comprises concatenating a plurality of three-dimensional (3D) tensors defined by the first spatial dimension, second spatial dimension, and input channel dimension to form the four-dimensional spatial input tensor; the combined batch index-temporal dimension has a size equal to the size of the batch index dimension of the input tensor multiplied by the size of the temporal dimension of the input tensor — This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). The claim describes the mathematical operations of concatenating tensors in words.
generating the 4D temporal input tensor comprises concatenating a plurality of 3D tensors defined by the batch index dimension, temporal dimension, and input channel dimension of the input tensor to form the four-dimensional temporal input tensor; the combined spatial dimension has a size equal to the size of the first spatial dimension of the input tensor multiplied by the size of the second spatial dimension of the input tensor; — This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). The claim describes the mathematical operations of concatenating tensors in words.
performing 2D convolution on the 4D spatial input tensor to generate the 4D spatial feature tensor comprises: applying a first number of 2D convolution filters to the 4D spatial input tensor, at a stride equal to a second number with respect to the first spatial dimension and the second spatial dimension, to generate the 4D spatial feature tensor having: a combined batch index-temporal output dimension; a first spatial output dimension, having a size equal to the first spatial dimension of the input tensor divided by the second number; a second spatial output dimension, having a size equal to the second spatial dimension of the input tensor divided by the second number; and an output channel dimension indicating a number of spatial feature maps equal to the first number; — This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). The claim describes the mathematical operations of performing tensor convolutions.
performing 2D convolution on the 4D temporal input tensor to generate the 4D temporal feature tensor comprises: applying a third number of 2D convolution filters to the 4D temporal input tensor, at a stride equal to a fourth number with respect to the temporal dimension, and a stride equal to the square of the fourth number with respect to the combined spatial dimension, to generate the 4D temporal feature tensor having: a batch index output dimension; a temporal dimension, having a size equal to the temporal dimension of the input tensor divided by the fourth number; a combined spatial output dimension, having a size equal to the first spatial dimension of the input tensor, divided by the fourth number, multiplied by the second spatial dimension of the input tensor, divided by the fourth number; and an output channel dimension indicating a number of temporal feature maps equal to the third number; — This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). The claim describes the mathematical operations of performing tensor convolutions.
processing the 4D spatial feature tensor and the 4D temporal feature tensor to generate the output tensor comprises: reshaping the 4D spatial feature tensor to generate a reshaped 4D spatial feature tensor, having: a batch index dimension having a size equal to the size of the batch index dimension of the input tensor; a temporal dimension having a size equal to the size of the temporal dimension of the input tensor; a combined spatial output dimension, having a size equal to: the first spatial dimension of the input tensor, divided by the second number, multiplied by the second spatial dimension of the input tensor, divided by the second number; and an output channel dimension indicating a number of spatial feature maps equal to the first number; — This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). The claim describes the mathematical operations of rearranging numbers into a different format and performing multiplications and divisions.
applying a number of single-element convolution filters equal to the first number to the reshaped 4D spatial feature tensor, at a temporal dimension stride equal to the fourth number with respect to the temporal dimension, to generate a temporally down-sampled 4D spatial feature tensor; and adding the elements of the temporally down-sampled 4D spatial feature tensor to the respective elements of the 4D temporal feature tensor to generate the output tensor; and the first number is equal to the third number — This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). The claim describes the mathematical operations of performing tensor convolutions.
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B.
Regarding Claim 10:
Step 1 – Is the claim to a process, machine, manufacture, or composition of matter?
Yes, the claim is to a machine.
Step 2A – Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
Yes, the claim recites the abstract ideas of:
generate a four-dimensional (4D) spatial input tensor based on the input tensor by combining the batch index dimension and temporal dimension of the input tensor into a combined batch index-temporal dimension; — This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). The claim limitations can be performed via tensor concatenation which is one kind of mathematical calculation.
perform two-dimensional (2D) convolution on the 4D spatial input tensor to generate a 4D spatial feature tensor; — This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). The claim describes the mathematical operations of performing convolutions on tensors.
generate a 4D temporal input tensor, based on the input tensor, having a combined spatial dimension based on the first spatial dimension and second spatial dimension of the input tensor; — This limitation is directed to the abstract idea of a mental process (including an observation, evaluation, judgement, opinion) which can be performed by the human mind, or by a human using pen and paper (see MPEP 2106.04(a)(2) III. C.).
perform 2D convolution on the 4D temporal input tensor to generate a 4D temporal feature tensor; — This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). The claim describes the mathematical operations of performing convolutions on tensors.
and process the 4D spatial feature tensor and the 4D temporal feature tensor to generate the output tensor, the output tensor comprising a 5D tensor — This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). According to figure 3 box 340, the output tensor can be generated by “addition and postprocessing”. Moreover, paragraph [0032] of the specification recites “ In some embodiments, the 4D temporal feature tensor 338 and the temporally down-sampled 4D spatial feature tensor 334 are combined through elementwise addition, resulting in an output tensor 129 of the same dimensions as the 4D temporal feature tensor 338 and the temporally down-sampled 4D spatial feature tensor 334”. Thus the claim limitation is directed to performing tensor calculations using elementwise addition.
Step 2A – Prong 2 – Does the claim recite additional elements that integrate the judicial exception into a practical application?
No, the claim does not recite additional elements that integrate the judicial exception into a practical application. The additional elements:
A system for processing an input tensor to generate an output tensor, the system comprising: a processor device; and a memory storing machine-executable instructions which, when executed by the processor device, cause the system to: — This limitation is directed to merely applying an abstract idea using a generic computer as a tool (see MPEP 2106.05(f)(2), 2106.04(d)).
obtain the input tensor, the input tensor comprising a five-dimensional (5D) tensor having: a batch index dimension of size B indicating individual video data samples in a batch of video data samples; a temporal dimension of size T; a first spatial dimension; a second spatial dimension; and an input channel dimension; — This limitation is directed to mere data gathering and outputting which has been recognized by the courts (as per Ultramercial, 772 F.3d at 715, 112 USPQ2d at 1754) as insignificant extra-solution activity (see MPEP 2106.05(g)).
Step 2B – Does the claim recite additional elements that amount to significantly more than the abstract idea itself?
No, the claim does not recite additional elements which amount to significantly more than the abstract idea itself. The additional elements as identified in step 2A prong 2:
A system for processing an input tensor to generate an output tensor, the system comprising: a processor device; and a memory storing machine-executable instructions which, when executed by the processor device, cause the system to: — Using a generic computer as a tool (see MPEP 2106.05(f)(2), 2106.05(d)) cannot amount to significantly more than the judicial exception itself.
obtain the input tensor, the input tensor comprising a five-dimensional (5D) tensor having: a batch index dimension of size B indicating individual video data samples in a batch of video data samples; a temporal dimension of size T; a first spatial dimension; a second spatial dimension; and an input channel dimension; — This limitation is recited at a high level of generality and amounts to mere data gathering of storing and retrieving information in memory, which is well-understood, routine, and conventional activity (see MPEP 2106.05(d) II.), which cannot amount to significantly more than the judicial exception.
Regarding Claim 11
Claim 11 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim is dependent on claim 10 which included an abstract idea (see rejection for claim 10). The claim recites the additional limitations:
Step 2A Prong 1:
wherein: generating the 4D temporal input tensor comprises concatenating a plurality of three-dimensional (3D) tensors defined by the first spatial dimension, second spatial dimension, and input channel dimension of the input tensor to form the 4D temporal input tensor; — This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). The claim describes the mathematical operations of tensor concatenation in words.
Step 2A Prong 2:
and the combined batch index-temporal dimension has a size equal to the size of the batch index dimension of the input tensor multiplied by the size of the temporal dimension of the input tensor — This limitation is directed to merely limiting a judicial exception to a particular field of use (see MPEP 2106.05(h)) as it merely limits the field of the combined batch index-temporal dimension.
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2.
Step 2B:
The additional elements as identified in step 2A prong 2:
and the combined batch index-temporal dimension has a size equal to the size of the batch index dimension of the input tensor multiplied by the size of the temporal dimension of the input tensor — Merely limiting a judicial exception to a particular field of use (see MPEP 2106.05(h)) cannot amount to significantly more than the judicial exception.
Thus, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B.
Regarding Claim 12
Claim 12 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim is dependent on claim 10 which included an abstract idea (see rejection for claim 10). The claim recites the additional limitations:
Step 2A Prong 1:
wherein: combining the first spatial dimension and second spatial dimension of the input tensor comprises concatenating a plurality of 3D tensors defined by the batch index dimension, temporal dimension, and input channel dimension to form the 4D temporal input tensor; — This limitation is directed to the abstract idea of a mathematical process, and mathematical calculations in particular (MPEP 2106.04(a)(2) I. C.). The claim describes the mathematical operations of tensor concatenation in words.
Step 2A Prong 2:
and the combined spatial dimension has a size equal to the size of the first spatial dimension of the input tensor multiplied by the size of the second spatial dimension of the input tensor — This limitation is directed to merely limiting a judicial exception to a particular field of use (see MPEP 2106.05(h)) as it merely limits the field of the combined spatial dimension.
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2.
Step 2B:
The additional elements as identified in step 2A prong 2:
and the combined spatial dimension has a size equal to the size of the first spatial dimension of the input tensor multiplied by the size of the second spatial dimension of the input tensor — Merely limiting a judicial exception to a particular field of use (see MPEP 2106.05(h)) cannot amount to significantly more than the judicial exception.
Thus, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B.
Regarding Claim 13
Claim 13 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim is dependent on claim 10 which included an abstract idea (see rejection for claim 13). The claim recites the additional limitations:
Step 2A Prong 1:
wherein performing 2D convolution on the 4D spatial input tensor to generate the 4D spatial feature tensor comprises: applying a first number of 2D convolution filters to the 4D spatial input tensor, at a stride equal to a second number with respect to the first spatial dimension and the second spatial dimension, to generate the 4D spatial feature tensor having: a combined batch index-temporal output dimension; a first spatial output dimension, having a size equal to the first spatial dimension of the input tensor divided by the second number; a second spatial output dimension, having a size equal to the second spatial dimension of the input tensor divided by the second number; and an output channel dimension indicating a number of spatial feature maps