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 .
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 10/20/2023 and 10/06/2025 are in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Status of Claims
Claim(s) 1, 9 and 17 is/are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Connah (NPL: Anysize GAN: A solution to the image-warping problem).
Claim(s) 2-5, 8, 10-13, 16 and 18-20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Connah (NPL: Anysize GAN: A solution to the image-warping problem) in view of Lin (US 20250054115 A1).
Claim(s) 6, 7, 14 and 15 is/are rejected under 35 U.S.C. 103 as being unpatentable over Connah (NPL: Anysize GAN: A solution to the image-warping problem) in view of Lin (US 20250054115 A1) further in view of Xue (NPL: Hiding Private Information in Images From AI).
Response to Arguments
Applicant's arguments filed 12/23/2025 have been fully considered but they are not persuasive. The applicant argues that “Even assuming, arguendo, that Connah et al. discloses, "learning the content of a single image and creating alternative representations of that image" and also discloses, "aspect ratio preserving resizing" it does not necessarily disclose, as Applicant has claimed, "the resizing enhances a prominence of the subject-of-interest within the captured image by changing an aspect ratio of the captured image to make the subject-of- interest more prominent;...”.
The examiner respectfully disagrees. Connah shows this in Fig. 1 wherein the images aspect ratio is normalized into a square shape. In the images, the subjects (the cheerleader, alligator, lizard, and dog) are stretched and have a larger focus within the image compared to their previous aspect ratios. In the broadest reasonable interpretation, the claim language “changing an aspect ratio of the captured image to make the subject-of- interest more prominent” is not indicative of how the subject is more prominent only that simply with an aspect ratio change the subject is more noticeable, visible, and focused. Without objective measurement of “prominence” of the subject or how aspect ratio changes are dictated, the figures of Connah teaches the embodiments of claim 1.
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 following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claim(s) 1, 9 and 17 is/are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Connah (NPL: Anysize GAN: A solution to the image-warping problem).
Regarding Claim 1:
Connah teaches: A computer implemented method for resizing a captured image, comprising (Page 6, “Our layer takes the desired output size, allowing the layer to resize to the users requirement. This is a unique layer differing to how other GANs perform resizing. For our research, we tested if a single resize layer would work or a progressively growing”):
receiving a desired size and a subject-of-interest the captured image as input from a user (Page 6, “Our layer takes the desired output size, allowing the layer to resize to the users requirement.”);
automatically resizing the captured image using a generative adversarial network (GAN) to about the desired size, wherein the resizing enhances a prominence of the subject-of-interest captured image by changing the aspect ratio of the captured image to make the subject-of-interest more prominent; (Page 6 and Figure 1, “Furthermore, SinGAN exceeds at learning the content of a single image and creating alternative representations of that image. However, it focuses on single images only matching content of that image i.e. on a dog the breed would always be the same, it cannot learn the differing breeds from a single image. In other words, the generated images are constraint by the context of the single trained image, and is unable to generate diverse features, e.g. for a skin lesion it will only produce samples of that image. Whereas, ours can produce unique images to expand datasets, like traditional GANs, such as DCGAN.”. In figure 1, pictures aspect ratios are normalized into a square shape giving more focus/prominence of the subject in the image); and
storing the automatically resized image on a computer readable storage medium (Page 8, “The network was trained using Tensorflow [23] with Keras [24] as the foreground API. Due to memory limitations, we employed an image aspect ratio preserving resizing.”).
Regarding Claim 9:
Connah teaches: A system, comprising:
an image sensor (Page 2, “However, most cameras use alternative aspect ratios, such as 16:9 in television, which means almost half the horizontal data would be lost in resizing”);
one or more processors (Page 8, “For our experiment, we used a machine with an RTX 2080 Ti (11GB) GPU, 128GB RAM and an Intel i7-7820x CPU on Windows 10. We used python 3.6 with Tensorflow version 1.13.1 and Keras 2.2.4 to design and run the models.); and
a memory communicatively coupled to the one or more processors (Page 8, “For our experiment, we used a machine with an RTX 2080 Ti (11GB) GPU, 128GB RAM and an Intel i7-7820x CPU on Windows 10. We used python 3.6 with Tensorflow version 1.13.1 and Keras 2.2.4 to design and run the models.”);
wherein the memory comprises instructions which, when executed by the one or more processors, cause the one or more processors to perform a method for resizing a captured image, comprising (Page 8, “For our experiment, we used a machine with an RTX 2080 Ti (11GB) GPU, 128GB RAM and an Intel i7-7820x CPU on Windows 10. We used python 3.6 with Tensorflow version 1.13.1 and Keras 2.2.4 to design and run the models.”):
receiving a desired size and a subject-of-interest the captured image as input from a user (Page 6, “Our layer takes the desired output size, allowing the layer to resize to the users requirement. This is a unique layer differing to how other GANs perform resizing. For our research, we tested if a single resize layer would work or a progressively growing”);
automatically resizing the captured image using a generative adversarial network (GAN) to about the desired size, wherein the resizing enhances a prominence of the subject-of-interest captured image by changing the aspect ratio of the captured image to make the subject-of-interest more prominent; (Page 6, “Furthermore, SinGAN exceeds at learning the content of a single image and creating alternative representations of that image. However, it focuses on single images only matching content of that image i.e. on a dog the breed would always be the same, it cannot learn the differing breeds from a single image. In other words, the generated images are constraint by the context of the single trained image, and is unable to generate diverse features, e.g. for a skin lesion it will only produce samples of that image. Whereas, ours can produce unique images to expand datasets, like traditional GANs, such as DCGAN.”); and
storing the automatically resized image on a computer readable storage medium (Page 8, “The network was trained using Tensorflow [23] with Keras [24] as the foreground API. Due to memory limitations, we employed an image aspect ratio preserving resizing.”).
Regarding Claim 17:
Connah teaches: A computer program product for resizing a captured image, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a computer system to perform a method for resizing a captured image, comprising (Page 6, “Our layer takes the desired output size, allowing the layer to resize to the users requirement. This is a unique layer differing to how other GANs perform resizing. For our research, we tested if a single resize layer would work or a progressively growing”):
receiving a desired size and a subject-of-interest within the captured image as input from a user (Page 6, “Our layer takes the desired output size, allowing the layer to resize to the users requirement.”);
automatically resizing the captured image using a generative adversarial network (GAN) to about the desired size, wherein the resizing enhances a prominence of the subject-of-interest captured image by changing the aspect ratio of the captured image to make the subject-of-interest more prominent; (Page 6, “Furthermore, SinGAN exceeds at learning the content of a single image and creating alternative representations of that image. However, it focuses on single images only matching content of that image i.e. on a dog the breed would always be the same, it cannot learn the differing breeds from a single image. In other words, the generated images are constraint by the context of the single trained image, and is unable to generate diverse features, e.g. for a skin lesion it will only produce samples of that image. Whereas, ours can produce unique images to expand datasets, like traditional GANs, such as DCGAN.”); and
storing the automatically resized image on the computer readable storage medium (Page 8, “The network was trained using Tensorflow [23] with Keras [24] as the foreground API. Due to memory limitations, we employed an image aspect ratio preserving resizing.”).
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status. The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim(s) 2-5, 8, 10-13, 16 and 18-20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Connah (NPL: Anysize GAN: A solution to the image-warping problem) in view of Lin (US 20250054115 A1).
Regarding Claim 2, 10 and 18:
Connah teaches the embodiments of claim 1, 9 and 17 as applied above.
Connah does not explicitly teach the following; however, in related art, Lin teaches: analyzing the image using the GAN to identify the subject and one or more other objects in the captured image (Paragraph 31-32, Paragraph 31 talks about identifying particular objects, such as a street sign, and in Paragraph 32 describes identifying multiple objects).
Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Lin’s multiple object detection GAN with Connah’s detection system using a GAN within an image.
Regarding Claim 3, 11 and 19:
Connah and Lin teach the embodiments of claim 2, 10 and 18 as applied above.
Lin further teaches: wherein the analyzing further comprises identifying the subject as a real and the one or more other objects as a fake (Paragraph 71, Objects are labeled as real or fake by image inpainting generator, a part of the GAN network described in Paragraph 69).
Regarding Claim 4, 12 and 20:
Connah and Lin teach the embodiments of claim 3, 11 and 19 as applied above.
Lin further teaches: wherein the enhancing comprises removing objects identified as fake (Paragraph 71, Fake objects are identified and then a mask is generated, this mask is then manipulated to blend the fake object within the image as a form of removing it.).
Regarding Claim 5 and 13:
Connah and Lin teach the embodiments of claim 3 and 11 as applied above.
Lin further teaches: wherein the automatic resizing comprises: adding additional fake input to the image by a generator component of the GAN; and removing the additional fake input by a discriminator component of the GAN (Paragraph 69-70, Inpainting generator is used to create fake inputs within the GAN where in the inpainting discriminator identifies the fake parts of an image to be removed).
Regarding Claim 8 and 16:
Connah teaches the embodiments of claim 1, 9 and 11 as applied above.
Connah does not explicitly teach the following; however, in related art, Lin further teaches: generating a higher resolution version of the identified subject using a Super Resolution GAN; and
replacing the identified subject with the generated higher resolution version of the main subject (Paragraph 105, multiple parts of multiple images are reduced or increased in resolution. Once the particular area of interest is identified, specific portions are replaced with higher resolution areas).
Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Lin to generate multiple resolutions of higher or lower scale on multiple parts of an image with Connah’s GAN system that can lower or increase the resolution of an entire image.
Claim(s) 6, 7, 14 and 15 is/are rejected under 35 U.S.C. 103 as being unpatentable over Connah (NPL: Anysize GAN: A solution to the image-warping problem) in view of Lin (US 20250054115 A1) further in view of Xue (NPL: Hiding Private Information in Images From AI).
Regarding Claim 6 and 14:
Connah and Lin teach the embodiments of claim 3 and 11 as applied above.
Connah and Lin do not explicitly teach the following; however, in related art, Xue teaches: wherein the analyzing further comprises identifying sensitive information in the captured image as fake (Page 4, “As shown in Fig. 5, the Blur and Mosaic’s “thickness” has been carefully adjusted to just hide the sensitive information”).
Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Xue to have sensitive information as an identifiable object in Connah and Lin’s GAN based object detection and image manipulation systems.
Regarding Claim 7 and 15:
Connah and Lin teach the embodiments of claim 6 and 14 as applied above.
Connah and Lin do not explicitly teach the following; however, in related art, Xue teaches: wherein the automatic resizing comprises blurring one or more other objects detected as fake (Page 4, “As shown in Fig. 5, the Blur and Mosaic’s “thickness” has been carefully adjusted to just hide the sensitive information”).
Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Xue’s blurring method on sensitive information with Connah’s and Lin’s GAN networks of identifying and manipulating objects within images.
Relevant Art directed to State of Art
Saharia (US 11978141 B2)
Narayan (US 10992902 B2)
MOHANTY (US 20240153306 A1)
DING (US 20240135613 A1)
Conclusion
THIS ACTION IS MADE FINAL. 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 DYLAN J SHERRILLO whose telephone number is (703)756-5605. The examiner can normally be reached 1st week of bi-week: Mon-Wed 7am-5:30pm PST, Thurs: 7am-4:30pm PST, Fri off / 2nd week of bi-week: Mon-Wed 7am-5:30pm PST, Thurs-Fri: 7am-4:30pm PST.
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, Stephen R Koziol can be reached at (408) 918-7630. 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.
/D.J.S./Examiner, Art Unit 2665
/Stephen R Koziol/Supervisory Patent Examiner, Art Unit 2665
Prosecution Insights