VIDEO FRAME FEATURE EXTRACTION METHOD, DEVICE AND COMPUTER-READABLE STORAGE MEDIUM

§102 §103 §112

DETAILED ACTION Notice of Pre-AIA or AIA Status The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA . Response to Amendment This action is responsive to the amendments and remarks received 24 June 2026. Claims 1 - 20 are currently pending. Claim Objections Claim 1 is objected to because of the following informalities: Line 4 of claim 1 recites, in part, “wherein each of the plurality of initial features is” which appears to contain a minor informality. The Examiner suggests amending the claim to --wherein each initial feature of the plurality of initial features is-- in order to improve the clarity and precision of the claim. Appropriate correction is required. Claim 7 is objected to because of the following informalities: Lines 14 - 15 of claim 7 recite, in part, “wherein the two-dimensional features of the initial feature of the target video frame has a height” which appears to contain a grammatical error and/or a minor informality. The Examiner suggests amending the claim to --wherein the two-dimensional features of the initial feature of the target video frame [[has]] have a height-- in order to improve the clarity and precision of the claim. Appropriate correction is required. Claim 8 is objected to because of the following informalities: Line 6 of claim 8 recites, in part, “wherein each of the plurality of initial features is” which appears to contain a minor informality. The Examiner suggests amending the claim to --wherein each initial feature of the plurality of initial features is-- in order to improve the clarity and precision of the claim. Appropriate correction is required. Claim 15 is objected to because of the following informalities: Line 5 of claim 15 recites, in part, “wherein each of the plurality of initial features is” which appears to contain a minor informality. The Examiner suggests amending the claim to --wherein each initial feature of the plurality of initial features is-- in order to improve the clarity and precision of the claim. Appropriate correction is required. The objections to claims 2 - 6, 9 - 14 and 16 - 20, due to minor informalities, are hereby withdrawn in view of the amendments and remarks received 24 June 2026. Claim Rejections - 35 USC § 112 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. Claim 7 is rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention. Claim 7 is rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention because it is unclear as to which initial feature of the target video frame “the initial feature of the target video frame” recited on lines 13 - 14, along with subsequent recitations of “the initial feature of the target video frame”, are referencing at least because lines 9 - 10 of claim 1, which recite, in part, “initial features of the target video frame;” (emphasis added) make it clear there are multiple initial features of the target video frame. Therefore, the Examiner asserts that it is unclear as to which initial feature of the multiple initial features of the target video frame “the initial feature of the target video frame” recited on lines 13 - 14 of claim 7, along with subsequent recitations of “the initial feature of the target video frame”, are referencing. Clarification and appropriate correction are required. For purposes of examination, the Examiner will treat the claim as referencing any initial feature of the multiple initial features of the target video frame. The rejections to claims 1 - 6 and 8 - 20 under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, are hereby withdrawn in view of the amendments and remarks received 24 June 2026. Response to Arguments Applicant’s arguments with respect to claim(s) 1 - 20 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument. Claim Rejections - 35 USC § 102 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 text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action. Claims 1 - 3, 5 - 10, 12 - 17, 19 and 20 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Luo et al. U.S. Publication No. 2022/0076002 A1. - With regards to claims 1, 8 and 15, Luo et al. disclose a computer-implemented method for extracting video frame features, (Luo et al., Abstract, Figs. 2, 3, 6A, 6B & 12 - 14, Pg. 1 ¶ 004 - 0008, Pg. 3 ¶ 0034 - 0041, Pg. 4 ¶ 0049 - 0050, Pg. 8 ¶ 0084, Pg. 16 ¶ 0182 - 0187) a device for extracting video frame features (Luo et al., Abstract, Figs. 2, 3, 6A, 6B & 12 - 14, Pg. 1 ¶ 004 - 0008, Pg. 3 ¶ 0034 - 0041, Pg. 4 ¶ 0049 - 0050, Pg. 8 ¶ 0084, Pg. 16 ¶ 0182 - 0187) comprising: one or more processors; (Luo et al., Fig. 14, Pg. 1 ¶ 0007 - 0008, Pg. 16 ¶ 0182 - Pg. 17 ¶ 0189) and a memory coupled to the one or more processors, the memory storing programs that, when executed by the one or more processors, cause performance of operations; (Luo et al., Fig. 14, Pg. 1 ¶ 0007 - 0008, Pg. 16 ¶ 0182 - Pg. 17 ¶ 0191) and a non-transitory computer-readable storage medium storing instructions that, when executed by at least one processor of a device, cause the at least one processor to perform a method, (Luo et al., Fig. 14, Pg. 1 ¶ 0007 - 0008, Pg. 16 ¶ 0182 - Pg. 17 ¶ 0191) the method/operations comprising: obtaining a plurality of initial features of each video frame of video frames in a video sequence, (Luo et al., Abstract, Figs. 3, 6A & 6B, Pg. 3 ¶ 0034 - 0041, Pg. 6 ¶ 0063, Pg. 7 ¶ 0071 and 0078, Pg. 8 ¶ 0084) wherein each of the plurality of initial features is a three-dimensional feature having a height dimension, a width dimension, and a feature channel dimension, (Luo et al., Fig. 6B, Pg. 3 ¶ 0034 - 0041, Pg. 7 ¶ 0078, Pg. 8 ¶ 0084) and wherein the feature channel dimension corresponds to a count of feature channels of the initial feature; (Luo et al., Fig. 6B, Pg. 3 ¶ 0034 - 0041, Pg. 7 ¶ 0076 - 0078, Pg. 8 ¶ 0084) calculating global channel attention information of the video sequence based on the plurality of initial features of each video frame in the video sequence; (Luo et al., Abstract, Figs. 3 & 5 - 6B, Pg. 3 ¶ 0034, 0037 and 0040 - 0042, Pg. 4 ¶ 0044 - 0049, Pg. 7 ¶ 0078, Pg. 8 ¶ 0084 [“After image data of video data in different temporal frames is obtained, the multi-channel convolutional layer is configured to obtain an original feature map of image data of each temporal frame, the original feature map including original feature submaps on different convolutional channels”]) calculating local channel attention information of a target video frame according to initial features of the target video frame; (Luo et al., Figs. 3 & 5 - 6B, Pg. 3 ¶ 0034, 0037 and 0040 - 0042, Pg. 4 ¶ 0044 - 0049, Pg. 7 ¶ 0078, Pg. 8 ¶ 0084) wherein the target video frame is one of the video frames in the video sequence; (Luo et al., Figs. 3 & 5 - 6B, Pg. 3 ¶ 0034, 0037 and 0040 - 0042, Pg. 4 ¶ 0044 - 0049, Pg. 7 ¶ 0078, Pg. 8 ¶ 0084) and performing channel attention mechanism processing on the initial features of the target video frame according to the global channel attention information and the local channel attention information to obtain optimized features of the target video frame. (Luo et al., Figs. 3 & 5 - 7, Pg. 3 ¶ 0034 and 0040 - 0042, Pg. 4 ¶ 0044 - 0050, Pg. 6 ¶ 0059 - 0060 and 0063 - 0066, Pg. 8 ¶ 0079 - 0084 [“Because the motion information weights are used for describing correlations between the original feature submaps of the image data of the target temporal frame on different convolutional channels and the action information of the moving object, by multiplying a motion information weight on each convolutional channel by an original feature submap on the corresponding channel to obtain motion information feature maps of the target temporal frame on the convolutional channels, and enhancing original feature submaps that are more relevant to the action information of the moving object and suppressing original feature submaps that are less relevant to the action information of the moving object, information that is beneficial to action recognition is enhanced and information that is irrelevant or even harmful to action recognition is suppressed, so that the motion information feature maps include more action information of the moving object, which is beneficial to subsequent action recognition on the moving object and effectively improves the accuracy of action recognition.”]) - With regards to claim 2, Luo et al. disclose the method of claim 1, wherein calculating the global channel attention information of the video sequence based on the plurality of initial features of each video frame in the video sequence comprises: performing fusion processing by averaging the plurality of initial features of each video frame in the video sequence to obtain global initial features of the video sequence, (Luo et al., Figs. 6A & 6B, Pg. 3 ¶ 0034 and 0040 - 0042, Pg. 7 ¶ 0071 - 0078, Pg. 8 ¶ 0084) wherein each feature channel of the global initial features corresponds to a two-dimensional feature having a height dimension and a width dimension; (Luo et al., Fig. 6B, Pg. 3 ¶ 0034 and 0041, Pg. 7 ¶ 0071 - 0078, Pg. 8 ¶ 0084) performing spatial dimension compression on the global initial features to obtain global initial channel attention information of the video sequence; (Luo et al., Figs. 6A & 6B, Pg. 3 ¶ 0034 and 0041, Pg. 7 ¶ 0071 - 0078, Pg. 8 ¶ 0084) and performing a purification operation on the global initial channel attention information to obtain the global channel attention information. (Luo et al., Figs. 6A & 6B, Figs. 6A & 6B, Pg. 3 ¶ 0034 and 0040 - 0042, Pg. 4 ¶ 0044 - 0049, Pg. 7 ¶ 0075 - Pg. 8 ¶ 0084) - With regards to claim 3, Luo et al. disclose the method of claim 2, wherein performing the spatial dimension compression on the global initial features to obtain the global initial channel attention information of the video sequence comprises: obtaining two-dimensional features corresponding to each feature channel in the global initial features; (Luo et al., Figs. 3, 6A & 6B, Pg. 3 ¶ 0034, 0037 and 0040 - 0042, Pg. 7 ¶ 0071 - 0078, Pg. 8 ¶ 0084) and performing the spatial dimension compression on the two-dimensional features corresponding to each feature channel to obtain the global initial channel attention information, (Luo et al., Figs. 6A & 6B, Pg. 3 ¶ 0034 and 0041, Pg. 7 ¶ 0071 - 0078, Pg. 8 ¶ 0084) wherein the spatial dimension compression comprises a global average pooling operation, (Luo et al., Figs. 6A & 6B, Pg. 7 ¶ 0071 - 0074 and 0078, Pg. 8 ¶ 0084) and wherein the global initial channel attention information has a height dimension, a width dimension, and a feature channel dimension, (Luo et al., Fig. 6B, Pg. 3 ¶ 0041, Pg. 7 ¶ 0072 - 0078, Pg. 8 ¶ 0084) and both the height dimension and the width dimension of the global initial channel attention information are equal to 1. (Luo et al., Fig. 6B, Pg. 3 ¶ 0041, Pg. 7 ¶ 0072 - 0078, Pg. 8 ¶ 0084) - With regards to claim 5, Luo et al. disclose the method of claim 1, wherein calculating local channel attention information of the target video frame according to the initial features of the target video frame comprises: performing spatial dimension compression on the initial features of the target video frame to obtain local initial channel attention information of the target video frame; (Luo et al., Figs. 6A & 6B, Pg. 3 ¶ 0034 and 0041, Pg. 7 ¶ 0071 - 0078, Pg. 8 ¶ 0084) and performing a purification operation on the local initial channel attention information to obtain the local channel attention information. (Luo et al., Figs. 6A & 6B, Figs. 6A & 6B, Pg. 3 ¶ 0034 and 0040 - 0042, Pg. 4 ¶ 0044 - 0049, Pg. 7 ¶ 0075 - Pg. 8 ¶ 0084) - With regards to claim 6, Luo et al. disclose the method of claim 1, wherein performing the channel attention mechanism processing on the initial features of the target video frame according to the global channel attention information and the local channel attention information to obtain the optimized features of the target video frame comprises: performing fusion processing on the global channel attention information and the local channel attention information to obtain fused channel attention information of the target video frame; (Luo et al., Figs. 3 & 5 - 7, Pg. 3 ¶ 0034 and 0040 - 0042, Pg. 4 ¶ 0044 - 0050, Pg. 6 ¶ 0059 - 0060 and 0063 - 0066, Pg. 8 ¶ 0079 - 0084) and performing the channel attention mechanism processing on the initial features of the target video frame to obtain the optimized features of the target video frame according to the fused channel attention information, (Luo et al., Figs. 3 & 6B, Pg. 4 ¶ 0044 - 0050, Pg. 8 ¶ 0079 - 0084 [“difference information corresponding to each convolutional channel is mapped into a motion information weight with a data value of 0 to 1 by using a sigmoid function. Subsequently, the motion information weight of each convolutional channel is multiplied by an original feature submap of the corresponding convolutional channel, so that feature information of original feature submaps of some convolutional channels is enhanced to different degrees, and feature information of original feature submaps of the remaining convolutional channels are suppressed to different degrees, to enhance feature information relevant to motion information in the original feature submaps of the target temporal frame by using feature information of the next temporal frame”]) wherein the fused channel attention information and the initial features of the target video frame are multiplied on a feature channel-by-feature channel basis to obtain the optimized features of the target video frame. (Luo et al., Figs. 3 & 6B, Pg. 4 ¶ 0044 - 0050, Pg. 8 ¶ 0079 - 0084 [“difference information corresponding to each convolutional channel is mapped into a motion information weight with a data value of 0 to 1 by using a sigmoid function. Subsequently, the motion information weight of each convolutional channel is multiplied by an original feature submap of the corresponding convolutional channel, so that feature information of original feature submaps of some convolutional channels is enhanced to different degrees, and feature information of original feature submaps of the remaining convolutional channels are suppressed to different degrees, to enhance feature information relevant to motion information in the original feature submaps of the target temporal frame by using feature information of the next temporal frame.”]) - With regards to claim 7, Luo et al. disclose the method of claim 6, wherein performing the channel attention mechanism processing on the initial features of the target video frame to obtain the optimized features of the target video frame according to the fused channel attention information comprises: obtaining channel attention information corresponding to each feature channel in the fused channel attention information, (Luo et al., Figs. 3 & 6B, Pg. 4 ¶ 0044 - 0050, Pg. 8 ¶ 0079 - 0084) wherein the channel attention information corresponding to each feature channel has a height dimension and a width dimension, and both the height dimension and the width dimension of the channel attention information corresponding to each feature channel are equal to 1; (Luo et al., Figs. 3 & 6B, Pg. 4 ¶ 0044 - 0050, Pg. 8 ¶ 0079 - 0084) and performing the channel attention mechanism processing on the initial features of the target video frame using the channel attention information corresponding to each feature channel to obtain the optimized features of the target video frame, (Luo et al., Figs. 3 & 6B, Pg. 4 ¶ 0044 - 0050, Pg. 8 ¶ 0079 - 0084 [“difference information corresponding to each convolutional channel is mapped into a motion information weight with a data value of 0 to 1 by using a sigmoid function. Subsequently, the motion information weight of each convolutional channel is multiplied by an original feature submap of the corresponding convolutional channel, so that feature information of original feature submaps of some convolutional channels is enhanced to different degrees, and feature information of original feature submaps of the remaining convolutional channels are suppressed to different degrees, to enhance feature information relevant to motion information in the original feature submaps of the target temporal frame by using feature information of the next temporal frame”]) wherein the channel attention information corresponding to each feature channel is used to perform the channel attention mechanism processing on two-dimensional features of the initial feature of the target video frame on the corresponding feature channel, (Luo et al., Figs. 3 & 6B, Pg. 3 ¶ 0034 and 0041, Pg. 4 ¶ 0044 - 0050, Pg. 7 ¶ 0078, Pg. 8 ¶ 0082 - 0084) wherein the two-dimensional features of the initial feature of the target video frame has a height dimension and a width dimension; (Luo et al., Figs. 3 & 6B, Pg. 3 ¶ 0034 and 0041, Pg. 7 ¶ 0074 and 0078, Pg. 8 ¶ 0084) and wherein performing the fusion processing on the global channel attention information and the local channel attention information to obtain the fused channel attention information of the target video frame comprises: averaging the global channel attention information and the local channel attention information to obtain the fused channel attention information of the target video frame. (Luo et al., Figs. 3 & 5 - 7, Pg. 3 ¶ 0034 and 0040 - 0042, Pg. 4 ¶ 0044 - 0048, Pg. 7 ¶ 0071 - 0075, Pg. 7 ¶ 0078 - Pg. 8 ¶ 0084 [“transform, by using a unit pooling layer, the original feature submaps of the target temporal frame on the convolutional channels into unit feature submaps of the target temporal frame, and the original feature submaps of the next temporal frame on the convolutional channels into unit feature submaps of the temporal frame” and “The unit pooling layer refers to a pooling layer configured to perform dimension reduction on original feature submaps. Optionally, the unit pooling layer may include an average pooling layer, for example, a global average pooling layer.”]) - With regards to claims 9 and 16, Luo et al. disclose the device and non-transitory computer-readable storage medium of claims 8 and 15, respectively, wherein calculating the global channel attention information of the video sequence based on the plurality of initial features of each video frame in the video sequence comprises: performing fusion processing on the plurality of initial features of each video frame in the video sequence to obtain global initial features of the video sequence; (Luo et al., Figs. 6A & 6B, Pg. 3 ¶ 0034 and 0040 - 0042, Pg. 7 ¶ 0071 - 0078, Pg. 8 ¶ 0084) performing spatial dimension compression on the global initial features to obtain global initial channel attention information of the video sequence; (Luo et al., Figs. 6A & 6B, Pg. 3 ¶ 0034 and 0041, Pg. 7 ¶ 0071 - 0078, Pg. 8 ¶ 0084) and performing a purification operation on the global initial channel attention information to obtain the global channel attention information. (Luo et al., Figs. 6A & 6B, Figs. 6A & 6B, Pg. 3 ¶ 0034 and 0040 - 0042, Pg. 4 ¶ 0044 - 0049, Pg. 7 ¶ 0075 - Pg. 8 ¶ 0084) - With regards to claims 10 and 17, Luo et al. disclose the device and non-transitory computer-readable storage medium of claims 9 and 16, respectively, wherein performing the spatial dimension compression on the global initial features to obtain the global initial channel attention information of the video sequence comprises: obtaining two-dimensional features corresponding to each feature channel in the global initial features; (Luo et al., Figs. 3, 6A & 6B, Pg. 3 ¶ 0034, 0037 and 0040 - 0042, Pg. 7 ¶ 0071 - 0078, Pg. 8 ¶ 0084) and performing the spatial dimension compression on the two-dimensional features corresponding to each feature channel to obtain the global initial channel attention information. (Luo et al., Figs. 6A & 6B, Pg. 3 ¶ 0034 and 0041, Pg. 7 ¶ 0071 - 0078, Pg. 8 ¶ 0084) - With regards to claims 12 and 19, Luo et al. disclose the device and non-transitory computer-readable storage medium of claims 8 and 15, respectively, wherein calculating the local channel attention information of the target video frame according to the initial features of the target video frame comprises: performing spatial dimension compression on the initial features of the target video frame to obtain local initial channel attention information of the target video frame; (Luo et al., Figs. 6A & 6B, Pg. 3 ¶ 0034 and 0041, Pg. 7 ¶ 0071 - 0078, Pg. 8 ¶ 0084) and performing a purification operation on the local initial channel attention information to obtain the local channel attention information. (Luo et al., Figs. 6A & 6B, Figs. 6A & 6B, Pg. 3 ¶ 0034 and 0040 - 0042, Pg. 4 ¶ 0044 - 0049, Pg. 7 ¶ 0075 - Pg. 8 ¶ 0084) - With regards to claims 13 and 20, Luo et al. disclose the device and non-transitory computer-readable storage medium of claims 8 and 15, respectively, wherein performing the channel attention mechanism processing on the initial features of the target video frame according to the global channel attention information and the local channel attention information to obtain the optimized features of the target video frame comprises: performing fusion processing on the global channel attention information and the local channel attention information to obtain fused channel attention information of the target video frame; (Luo et al., Figs. 3 & 5 - 7, Pg. 3 ¶ 0034 and 0040 - 0042, Pg. 4 ¶ 0044 - 0050, Pg. 6 ¶ 0059 - 0060 and 0063 - 0066, Pg. 8 ¶ 0079 - 0084) and performing the channel attention mechanism processing on the initial features of the target video frame to obtain the optimized features of the target video frame according to the fused channel attention information. (Luo et al., Figs. 3 & 6B, Pg. 4 ¶ 0044 - 0050, Pg. 8 ¶ 0079 - 0084 [“difference information corresponding to each convolutional channel is mapped into a motion information weight with a data value of 0 to 1 by using a sigmoid function. Subsequently, the motion information weight of each convolutional channel is multiplied by an original feature submap of the corresponding convolutional channel, so that feature information of original feature submaps of some convolutional channels is enhanced to different degrees, and feature information of original feature submaps of the remaining convolutional channels are suppressed to different degrees, to enhance feature information relevant to motion information in the original feature submaps of the target temporal frame by using feature information of the next temporal frame.”]) - With regards to claim 14, Luo et al. disclose the device of claim 13, wherein performing the channel attention mechanism processing on the initial features of the target video frame to obtain the optimized features of the target video frame according to the fused channel attention information comprises: obtaining channel attention information corresponding to each feature channel in the fused channel attention information; (Luo et al., Figs. 3 & 6B, Pg. 4 ¶ 0044 - 0050, Pg. 8 ¶ 0079 - 0084) and performing the channel attention mechanism processing on the initial features of the target video frame using the channel attention information corresponding to each feature channel to obtain the optimized features of the target video frame. (Luo et al., Figs. 3 & 6B, Pg. 4 ¶ 0044 - 0050, Pg. 8 ¶ 0079 - 0084 [“difference information corresponding to each convolutional channel is mapped into a motion information weight with a data value of 0 to 1 by using a sigmoid function. Subsequently, the motion information weight of each convolutional channel is multiplied by an original feature submap of the corresponding convolutional channel, so that feature information of original feature submaps of some convolutional channels is enhanced to different degrees, and feature information of original feature submaps of the remaining convolutional channels are suppressed to different degrees, to enhance feature information relevant to motion information in the original feature submaps of the target temporal frame by using feature information of the next temporal frame.”]) Claim Rejections - 35 USC § 103 The text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action. Claims 4, 11 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Luo et al. U.S. Publication No. 2022/0076002 A1 as applied to claims 2, 9 and 16 above, and further in view of Jie Hu, Li Shen, Samuel Albanie, Gang Sun, and Enhua Wu, “Squeeze-and-Excitation Networks”, IEEE, IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 42, No. 8, Aug. 2020, pages 2011 - 2023, herein referred to as “Hu et al.”. - With regards to claim 4, Luo et al. disclose the method of claim 2, wherein performing the purification operation on the global initial channel attention information to obtain the global channel attention information comprises: performing dimensionality reduction processing on the global initial channel attention information using a preset first layer to obtain global dimensionality reduction channel attention information; (Luo et al., Figs. 6A & 6B, Figs. 6A & 6B, Pg. 3 ¶ 0034 and 0040 - 0042, Pg. 4 ¶ 0044 - 0049, Pg. 7 ¶ 0075 - Pg. 8 ¶ 0084) and performing dimensionality enhancement processing on the global dimensionality reduction channel attention information using a preset second layer to obtain second channel attention information, (Luo et al., Figs. 6A & 6B, Pg. 8 ¶ 0079 - 0084) and activating the second channel attention information using a preset second activation function to obtain the global channel attention information. (Luo et al., Figs. 6A & 6B, Pg. 7 ¶ 0068 - 0070, Pg. 8 ¶ 0079 - 0084, Pg. 11 ¶ 0111 - 0114) Luo et al. fail to disclose explicitly using a preset first fully connected layer to obtain first channel attention information, and activating the first channel attention information using a preset first activation function; and using a preset second fully connected layer. Pertaining to analogous art, Hu et al. disclose performing dimensionality reduction processing on the global initial channel attention information using a preset first fully connected layer to obtain first channel attention information, (Hu et al., Pg. 2011 § 1 ¶ 2 - 3, Pg. 2012 Fig. 1, Pg. 2013 § 3 - § 3.2, Pgs. 2014 - 2015 § 4, Pg. 2014 Figs. 2 - 3) and activating the first channel attention information using a preset first activation function to obtain global dimensionality reduction channel attention information; (Hu et al., Pg. 2011 § 1 ¶ 2 - 3, Pg. 2012 Fig. 1, Pg. 2013 § 3.2, Pg. 2014 Figs. 2 - 3, Pg. 2018 § 6.3) and performing dimensionality enhancement processing on the global dimensionality reduction channel attention information using a preset second fully connected layer to obtain second channel attention information, (Hu et al., Pg. 2011 § 1 ¶ 2 - 3, Pg. 2012 Fig. 1, Pg. 2013 § 3 - § 3.2, Pgs. 2014 - 2015 § 4, Pg. 2014 Figs. 2 - 3) and activating the second channel attention information using a preset second activation function to obtain the global channel attention information. (Hu et al., Pg. 2011 § 1 ¶ 2 - 3, Pg. 2012 Fig. 1, Pg. 2013 § 3.2, Pg. 2014 Figs. 2 - 3, Pg. 2018 § 6.3) Luo et al. and Hu et al. are combinable because they are both directed towards computer vision systems that aim to emphasize relevant and informative image features and suppress less useful image features. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings of Luo et al. with the teachings of Hu et al. This modification would have been prompted in order to substitute the dimensionality reduction and enhancement processing techniques of Luo et al. for the dimensionality reduction and increasing processes of Hu et al. The dimensionality reduction and increasing processes of Hu et al. could be substituted in place of the dimensionality reduction and enhancement processing techniques of Luo et al. utilizing well-known techniques in the art and would likely yield predictable results, in that, in the combination, the dimensionality reduction processing would be performed by the preset first fully-connected layer followed by ReLU (Rectified Linear Unit) activation function of Hu et al. and the dimensionality enhancement processing would be performed by the preset second fully-connected layer followed by sigmoid activation function of Hu et al. Furthermore, this modification would have been prompted by the teachings and suggestions of Hu et al. that their teachings improve the representational power of features, can be integrated into standard architectures such as ResNet-50 and may prove useful for tasks requiring strong discriminative features, see at least page 2011 section 1 paragraphs 2 - 3, page 2013 section 3, page 2013 section 3.3 - page 2014 section 4 paragraph 1 and page 2021 section 8 of Hu et al. This combination could be completed according to well-known techniques in the art and would likely yield predictable results, in that the preset first fully-connected layer followed by ReLU (Rectified Linear Unit) activation function and the dimensionality enhancement processing would be performed by the preset second fully-connected layer followed by sigmoid activation function of Hu et al. would be utilized to perform the dimensionality reduction processing and the dimensionality enhancement processing, respectively, of the base device of Luo et al. Therefore, it would have been obvious to combine Luo et al. with Hu et al. to obtain the invention as specified in claim 4. - With regards to claims 11 and 18, Luo et al. disclose the device and non-transitory computer-readable storage medium of claims 9 and 16, respectively, wherein performing the purification operation on the global initial channel attention information to obtain the global channel attention information comprises: performing dimensionality reduction processing on the global initial channel attention information using a preset first layer to obtain global dimensionality reduction channel attention information; (Luo et al., Figs. 6A & 6B, Figs. 6A & 6B, Pg. 3 ¶ 0034 and 0040 - 0042, Pg. 4 ¶ 0044 - 0049, Pg. 7 ¶ 0075 - Pg. 8 ¶ 0084) and performing dimensionality enhancement processing on the global dimensionality reduction channel attention information using a preset connected layer to obtain the global channel attention information. (Luo et al., Figs. 6A & 6B, Pg. 7 ¶ 0068 - 0070, Pg. 8 ¶ 0079 - 0084, Pg. 11 ¶ 0111 - 0114) Luo et al. fail to disclose explicitly using a preset first fully connected layer; and using a preset second fully connected layer. Pertaining to analogous art, Hu et al. disclose performing dimensionality reduction processing on the global initial channel attention information using a preset first fully connected layer to obtain global dimensionality reduction channel attention information; (Hu et al., Pg. 2011 § 1 ¶ 2 - 3, Pg. 2012 Fig. 1, Pg. 2013 § 3 - § 3.2, Pgs. 2014 - 2015 § 4, Pg. 2014 Figs. 2 - 3) and performing dimensionality enhancement processing on the global dimensionality reduction channel attention information using a preset second fully connected layer to obtain the global channel attention information. (Hu et al., Pg. 2011 § 1 ¶ 2 - 3, Pg. 2012 Fig. 1, Pg. 2013 § 3 - § 3.2, Pgs. 2014 - 2015 § 4, Pg. 2014 Figs. 2 - 3) Luo et al. and Hu et al. are combinable because they are both directed towards computer vision systems that aim to emphasize relevant and informative image features and suppress less useful image features. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings of Luo et al. with the teachings of Hu et al. This modification would have been prompted in order to substitute the dimensionality reduction and enhancement processing techniques of Luo et al. for the dimensionality reduction and increasing processes of Hu et al. The dimensionality reduction and increasing processes of Hu et al. could be substituted in place of the dimensionality reduction and enhancement processing techniques of Luo et al. utilizing well-known techniques in the art and would likely yield predictable results, in that, in the combination, the dimensionality reduction processing would be performed by the preset first fully-connected layer followed by ReLU (Rectified Linear Unit) activation function of Hu et al. and the dimensionality enhancement processing would be performed by the preset second fully-connected layer followed by sigmoid activation function of Hu et al. Furthermore, this modification would have been prompted by the teachings and suggestions of Hu et al. that their teachings improve the representational power of features, can be integrated into standard architectures such as ResNet-50 and may prove useful for tasks requiring strong discriminative features, see at least page 2011 section 1 paragraphs 2 - 3, page 2013 section 3, page 2013 section 3.3 - page 2014 section 4 paragraph 1 and page 2021 section 8 of Hu et al. This combination could be completed according to well-known techniques in the art and would likely yield predictable results, in that the preset first fully-connected layer followed by ReLU (Rectified Linear Unit) activation function and the dimensionality enhancement processing would be performed by the preset second fully-connected layer followed by sigmoid activation function of Hu et al. would be utilized to perform the dimensionality reduction processing and the dimensionality enhancement processing, respectively, of the base device of Luo et al. Therefore, it would have been obvious to combine Luo et al. with Hu et al. to obtain the invention as specified in claims 11 and 18. Conclusion Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a). A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action. Any inquiry concerning this communication or earlier communications from the examiner should be directed to ERIC RUSH whose telephone number is (571) 270-3017. The examiner can normally be reached 9am - 5pm Monday - Friday. 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, Andrew Bee can be reached at (571) 270 - 5183. 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. /ERIC RUSH/Primary Examiner, Art Unit 2677