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 .
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention
was made.
Claim(s) 1 - 6, 8, 10 -14, and 17, 19-20 are rejected under 35 U.S.C. 103 as being unpatentable over BETHGE (US-20180158224-A1) in view of Nangi (US-20230137209-A1) .
Regarding claim 1, BETHGE teaches A computer-implemented method for performing style transfer, the method comprising (Para. 2 and 22: teaches a computer implemented image style transfer method); a first set of features associated with a content sample into a second set of features from a feature space associated with one or more style samples (Para.15, para.19 and fig 1a/1b: teaches the image extraction feature and the content features and style features); computing one or more losses based on the first set of features and the second set of features (Para.24 -25 and fig 3: teaches content loss based on content features ad style loss based on style features); and generating a style transfer result based on the content sample and the one or more losses(Para.23, 31-33 and fig.3: teaches generating a stylized image by minimizing losses via backpropagation), wherein the style transfer result comprises one or more content-based attributes of the content sample and one or more style-based attributes of the one or more style samples ( Para.20-22 and para 26: teaches output preserves content structure while applying style features).
BETHGE Fails to teach converting, via a trained variational autoencoder
Nangi Teaches converting, via a trained variational autoencoder (Para 42-43 and 60-61: teaches a trained VAE that takes input (content sample) and encodes into latent feature space. It also teaches separating content features 206 and style features 208. It would have been obvious to modify the image style transfer method of Bethge to incorporate the variational autoencoder-based feature conversion of Nangi in order to improve the style transfer framework to provide more efficient, structured, and controllable feature transformations.)
Regarding claim 2, BETHGE in view of Nangi teaches The computer-implemented method of claim 1, further comprising: converting, via a variational autoencoder, a third set of features associated with the one or more style samples into a fourth set of features (Bethge, Para.19-20 and fig.2b: teaches style features extracted from style samples . Nangi, Para.42-43, 60, and 75: teaches VAE encoding into latent style representation); computing one or more additional losses based on the third set of features and the fourth set of features (Bethge Para. 24-25: style loss based on feature based on feature correlations. Nangi, Para 61-67: teaches multiple additional losses specifically tied to style representations including adversarial style loss and multitask style loss); and generating the trained variational autoencoder by training the variational autoencoder based on the one or more additional losses (Nangi ,Para.46. 61 and 66-67: teaches training the VAE using multiple losses and losses directly influence latent representations) .
Regarding claim 3, BETHGE in view of Nangi teaches The computer-implemented method of claim 1, further comprising extracting the first set of features using a feature extractor neural network (Bethge, Para.14-15: teaching extracting feature maps from an image).
Regarding claim 4, BETHGE in view of Nangi teaches The computer-implemented method of claim 3, wherein the first set of features is extracted from a plurality of layers included in the feature extractor neural network ( BETHGE , Para 14-16 and 23-24: teaches extracting and using features from multiple CNN layers (L, a, b, c).) .
Regarding claim 5, BETHGE in view of Nangi teaches The computer-implemented method of claim 1, wherein generating the style transfer result comprises iteratively updating the content sample based on the one or more losses (BETHGE, Para.23, para.32-33: teaches repeatedly adjusting (updating) the image, i.e., iterative optimization. The “adjusted image” at each step = the updated content sample.).
Regarding claim 6, BETHGE in view of Nangi teaches The computer-implemented method of claim 1, wherein converting the first set of features into the second set of features comprises: converting, by an encoder neural network included in the trained variational autoencoder, the first set of features into one or more embeddings within an embedding space(Nangi, Para.42, 44, 46: teaches an encoder neural network that maps input features into a latent embedding space); and converting, by a decoder neural network included in the trained variational autoencoder, the one or more embeddings into the second set of features (para.45 and para 55: teaches a decoder neural network that converts latent embeddings back into output representations).
Regarding claim 8, BETHGE in view of Nangi teaches The computer-implemented method of claim 1, wherein the content sample and the one or more style samples comprise at least one of an image (Bethge, para 9 and fig 1a/1b: teaches transfer of a texture of a source image to a target image ).
Regarding claim 10, BETHGE in view of Nangi teaches The computer-implemented method of claim 1, wherein the one or more losses comprise a distance between the first set of features and the second set of features (BETHGE, Para. 25, 27-30: teaches computing a loss based of differences between feature representations including content and style feature maps).
Regarding claim 11, BETHGE teaches One or more non-transitory computer-readable media storing instructions that, when executed by one or more processors, cause the one or more processors to perform the steps of: (Para. 2 and 22: teaches a computer implemented image style transfer method); a first set of features associated with a content sample into a second set of features from a feature space associated with one or more style samples (Para.15, para.19 and fig 1a/1b: teaches the image extraction feature and the content features and style features); computing one or more losses based on the first set of features and the second set of features (Para.24 -25 and fig 3: teaches content loss based on content features ad style loss based on style features); and generating a style transfer result based on the content sample and the one or more losses(Para.23, 31-33 and fig.3: teaches generating a stylized image by minimizing losses via backpropagation), wherein the style transfer result comprises one or more content-based attributes of the content sample and one or more style-based attributes of the one or more style samples ( Para.20-22 and para 26: teaches output preserves content structure while applying style features).
BETHGE Fails to teach converting, via a trained variational autoencoder
Nangi Teaches converting, via a trained variational autoencoder (Para 42-43 and 60-61: teaches a trained VAE that takes input (content sample) and encodes into latent feature space. It also teaches separating content features 206 and style features 208. It would have been obvious to modify the image style transfer method of Bethge to incorporate the variational autoencoder-based feature conversion of Nangi in order to improve the style transfer framework to provide more efficient, structured, and controllable feature transformations.).
Regarding claim 12, It falls under the same rejection as claim 3 because it is similar in scope and dependent upon same references.
Regarding claim 13, It falls under the same rejection as claim 4 because it is similar in scope and dependent upon same references.
Regarding claim 14, BETHGE in view of Nangi teaches The one or more non-transitory computer-readable media of claim 13, wherein the trained variational autoencoder comprises a plurality of encoder-decoder pairs corresponding to the plurality of layers (Nangi, Para. 54-56:the encoder 308 and decoder 322 are both composed of multiple layers ).
Regarding claim 17, It falls under the same rejection as claim 6 because it is similar in scope and dependent upon same references.
Regarding claim 19, It falls under the same rejection as claim 10 because it is similar in scope and dependent upon same references.
Regarding claim 20, BETHGE teaches A system, comprising: one or more memories that store instructions, and one or more processors that are coupled to the one or more memories and, when executing the instructions, are configured to perform the steps of: (Para. 2 and 22: teaches a computer implemented image style transfer method); a first set of features associated with a content sample into a second set of features from a feature space associated with one or more style samples (Para.15, para.19 and fig 1a/1b: teaches the image extraction feature and the content features and style features); computing one or more losses based on the first set of features and the second set of features (Para.24 -25 and fig 3: teaches content loss based on content features ad style loss based on style features); and generating a style transfer result based on the content sample and the one or more losses(Para.23, 31-33 and fig.3: teaches generating a stylized image by minimizing losses via backpropagation), wherein the style transfer result comprises one or more content-based attributes of the content sample and one or more style-based attributes of the one or more style samples ( Para.20-22 and para 26: teaches output preserves content structure while applying style features).
BETHGE Fails to teach converting, via a trained variational autoencoder
Nangi Teaches converting, via a trained variational autoencoder (Para 42-43 and 60-61: teaches a trained VAE that takes input (content sample) and encodes into latent feature space. It also teaches separating content features 206 and style features 208. It would have been obvious to modify the image style transfer method of Bethge to incorporate the variational autoencoder-based feature conversion of Nangi in order to improve the style transfer framework to provide more efficient, structured, and controllable feature transformations.)
Claim(s) 7 are rejected under 35 U.S.C. 103 as being unpatentable over BETHGE (US-20180158224-A1) in view of Nangi (US-20230137209-A1) in further view of Chandran (Made of reference in IDS: US-20220156987-A1).
Regarding claim 7, BETHGE in view of Nangi teaches The computer-implemented method of claim 1, wherein the trained variational autoencoder but fails to teach comprises a first encoder-decoder pair associated with a first subset of the first set of features and a second encoder-decoder pair associated with a second subset of the first set of features.
Chandran teaches a first encoder-decoder pair associated with a first subset of the first set of features and a second encoder-decoder pair associated with a second subset of the first set of features ( Para.29-32: teaches multiple encoders each tied to different subset of features. Each encoder inherently operates with a corresponding decoding stage forming an encoder decoder pair. It would have been obvious to modify the system of BETHGE in view of Nangi to include multiple encoder-decoder pairs corresponding to different feature subsets, in order to improve feature disentanglement and representation accuracy).
Claim(s) 9 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over BETHGE (US-20180158224-A1) in view of Nangi (US-20230137209-A1) WEINMANN (US-20220215510-A1).
Regarding claim 9, BETHGE in view of Nangi teaches The computer-implemented method of claim 1, wherein the one or more losses are computed (Bethge, Para. 24-25 and fig.3: teaches loss functions and para 27-28: teaches loss computed between feature representations across layers.).
BETHGE in view of Nangi Fails to teach first set of normalized features corresponding to the first set of features and a second set of normalized features corresponding to the second set of features
WEINMANN teaches first set of normalized features corresponding to the first set of features and a second set of normalized features corresponding to the second set of features( para 158 and fig 1/s30: teaches normalized features corresponding to a first and second set and computing difference. It would have been obvious to normalize feature representations prior to computing differences as taught by WEINMANN , in order to ensure stable loss computation between feature representations).
Regarding claim 18, It falls under the same rejection as claim 9 because it is similar in scope and dependent upon same references.
Claim(s) 15 is rejected under 35 U.S.C. 103 as being unpatentable over BETHGE (US-20180158224-A1) in view of Nangi (US-20230137209-A1) in further view of Wang (US-20250265693-A1).
Regarding claim 15, BETHGE in view of Nangi teaches The one or more non-transitory computer-readable media of claim 11, wherein the instructions further cause the one or more processors to perform the steps of: converting, via the trained variational autoencoder, a third set of features associated with a sample into a fourth set of features (Bethge, Para.19-20 and fig.2b: teaches style features extracted from style samples . Nangi, Para.42-43, 60, and 75: teaches VAE encoding into latent style representation); computing one or more additional losses between the third set of features and the fourth set of features; and generating the style transfer result based on the one or more additional losses (Bethge Para. 24-25: style loss based on feature based on feature correlations. Nangi, Para 61-67: teaches multiple additional losses specifically tied to style representations including adversarial style loss and multitask style loss).
BETHGE in view of Nangi fail to teach a scaled version of the content sample. Wang teaches scaled version of content sample ( Para.17: teaches generating scaled versions of an input image at multiple sizes and extracting features. It would have been obvious to apply such scaling to the system of BETHGE in view of Nangi VAE to improve feature accuracy across scaling).
Claim(s) 16 are rejected under 35 U.S.C. 103 as being unpatentable over BETHGE (US-20180158224-A1) in view of Nangi (US-20230137209-A1) in further view of Hamalainen (US-20100202659-A1).
Regarding claim 16, BETHGE in view of Nangi teaches The one or more non-transitory computer-readable media of claim 14, but fails to teach wherein the instructions further cause the one or more processors to perform the step of sampling a scale associated with the scaled version of the content sample from a distribution.
Hamalainen teaches wherein the instructions further cause the one or more processors to perform the step of sampling a scale associated with the scaled version of the content sample from a distribution( Para.51: teaches the technique of sampling a scale associated with a scaled version. It would have been obvious to incorporate Hamalainen technique of sampling scale from a distribution into the system of BETHGE in view of Nangi in order to improves scaling across the samples ).
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to LATRELL ANTHONY CREARY whose telephone number is (703)756-1219. The examiner can normally be reached Mon - Fri 7:30am - 4:30pm.
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, Xiao WU can be reached on (571) 272-7761. 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.
/LATRELL ANTHONY CREARY/Examiner, Art Unit 2613
/XIAO M WU/Supervisory Patent Examiner, Art Unit 2613
Prosecution Insights