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 .
Election/Restrictions
Claims 10-14 are withdrawn from further consideration pursuant to 37 CFR 1.142(b), as being drawn to a nonelected invention, there being no allowable generic or linking claim. Applicant timely traversed the restriction (election) requirement in the reply filed on 04/20/2026.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 04/03/2024 is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
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.
Claims 1, 5, 7 are rejected under 35 U.S.C. 103 as being unpatentable over MEN et al. (“Controllable Person Image Synthesis with Attribute-Decomposed GAN”) in view of ZHOU et al. (“Cross Attention Based Style Distribution for Controllable Person Image Synthesis”).
RE claim 1, Men teaches a method that allows for the synthesis of more realistic output images and automatic separation of un-annotated attributes [0001]. Men teaches a method for image generation, comprising:
(a)
obtaining an input image depicting an entity and a skeleton map depicting a pose of the entity;
Fig. 2, source person image (Is) (said input image) [0013]. Men teaches the corresponding keypoint-based pose (P), 18 channel heat map that encodes the locations of 18 joints of a human body (said skeleton map), can be automatically extracted via existing pose estimation method [0012].
(b)
performing, using an image generation model, a cross-attention mechanism between image features of the input image and entity features representing the pose to obtain modified image features; and
Fig. 2 of Men illustrates the generator whose inputs are the target pose (Pt) and a source person image (Is), and the output is the generated image (Ig) with source person (Is) in the target pose (Pt) [0013]. The desired person image is reconstructed from target features by a decoder (Ig) [0013]. However, Men does not discuss image features.
Zhou teaches a cross attention based style distribution (CASD) module that computes between the source semantic styles and target pose for pose transfer [0001, 0004, 0006]. The module intentionally selects the style represented by each semantic and distributes them according to the target pose [0001]. Zhou teaches given a source image (Is) (said input image) under the pose (Ps) (said skeleton map), the system/method of Zhou synthesizes a high fidelity image (Ît) (said modified image) under a different target pose (Pt) [0013]. As shown in Fig. 2, the desired pose (Pt) is directly used as the input by encoder (Ep), which describes the key point positions of human body (said entity features) [0014]. To facilitate accurate style extractions (said image features from input image) from source image (Is), the method/system of Zhou employs source parsing map (Ss) to separate the full image into regions, so that Es independently encodes the styles in different semantics [0014]. As shown in Fig. 2, Fp represents pose features from Ep (said entity features) and Fs represents style features from Es (said image features) [0014]. They are utilized by the CASD blocks [0014]. The cross attention module adapts the source style into the required pose (F’s) [Fig. 3, 0020].
It would have been obvious before the effective filing date of the claimed invention to utilize the CASD model of Zhou with the image synthesis of Men because the attention matrix in cross attention expresses the dynamic similarities between the target pose and the source styles for all semantics. Therefore, it can be utilized to route the color and texture from the source image, and is further constrained by the target parsing map to achieve a clearer objective [Zhou: 0001].
(c)
generating, using the image generation model, an output image based on the modified image features, wherein the output image depicts the entity with the pose.
Fig. 2 of Men illustrates the generator whose inputs are the target pose (Pt) and a source person image (Is), and the output is the generated image (Ig) with source person (Is) in the target pose (Pt) (said generating output image depicts the entity with the pose) [0013]. The desired person image is reconstructed from target features by a decoder (Ig) [0013]. As further taught in the rationale of claim 1(b), Zhou is relied upon as teaching a CASD model. The cross attention module adapts the source style into the required pose (F’s) (said generating output image based on modified image features, depts entity with the pose) [Fig. 3, 0020].
RE claim 5, Men in view of Zhou teaches further comprising:
(a)
encoding the input image to obtain the image features; and
In further view of Zhou, Zhou teaches Zhou teaches given a source image (Is) (said input image) under the pose (Ps), the system/method of Zhou synthesizes a high fidelity image (Ît) under a different target pose (Pt) [0013]. As shown in Fig. 2, the desired pose (Pt) is directly used as the input by encoder (Ep), which describes the key point positions of human body (said entity features) [0014]. To facilitate accurate style extractions (said image features) from source image (Is), the method/system of Zhou employs source parsing map (Ss) to separate the full image into regions, so that Es independently encodes the styles in different semantics (said encoding input image to obtain image features) [0014].
(b)
encoding a first region of the input image surrounding the entity to obtain the entity features.
As taught in claim 5(a), a source parsing map is used to separate the full image into regions (said encoding first region of input image), so that Es independently encodes the styles in different semantics [0014].
The same motivation to combine as taught in the rationale of claim 1 is incorporated herein.
RE claim 7, in further view of Zhou, Zhou teaches wherein performing the cross-attention mechanism comprises:
(a)
computing a key vector and a value vector for the entity.
Zhou teaches using keys and values within the self-attention modules [0012]. The computing is also described in [0018-0021].
The same motivation to combine as taught in the rationale of claim 1 is incorporated herein.
Claims 2, 3, 6 are rejected under 35 U.S.C. 103 as being unpatentable over MEN et al. (“Controllable Person Image Synthesis with Attribute-Decomposed GAN”) in view of ZHOU et al. (“Cross Attention Based Style Distribution for Controllable Person Image Synthesis”) as applied to claim 1, and in further view of LIN et al. (US 2024/0127410 A1).
RE claim 2, Men in view of Zhou teaches the limitations of claim 2 with the exception of discussing inpainting. However, Lin is made of record as teaching a system/method for panoptically guiding digital image inpainting utilizing a panoptic inpainting neural network [abstract]. Lin teaches further comprising:
(a)
obtaining an inpainting mask indicating an interaction region of the entity with an additional entity,
With reference to Fig. 2 of Lin, the panoptic inpainting system (102) receives a request to generate an inpainted digital image. The panoptic inpainting system (102) identifies the designated area of missing or flawed pixels indicated by a binary mask (said inpainting mask) obfuscating or occluding a portion of the digital image depicting four women (said entity with an additional entity) against a mountain backdrop [0061].
(b)
wherein the output image is generated based on the inpainting mask.
Lin teaches the panoptic inpainting system (102) performs an act (208) to generate inpainted digital image [0064].
It would have been obvious before the effective filing date of the claimed invention of Men in view of Zhou to use the inpainting technology of Lin in order correct missing or flawed pixels, especially in the case of occluding areas of multiple objects, such as the four women [0061]. As Lin teaches, it is common in the cases where a designated area of a digital image depicts two objects that are adjacent to one another or that overlap by some portion, and further share a common semantic label, to be have the designated area be generated by one misshapen block of pixels that results from attempting to generate one object of the semantic label where the two separate objects should appear [Lin: 0036]. The method/system of Lin helps to maintain the boundary of the two objects depicted.
RE claim 3, Men in view of Zhou teaching the limitations of claim 3 with the exception of obscured interaction between multiple entities. However, Lin is made of record as teaching a system/method for panoptically guiding digital image inpainting utilizing a panoptic inpainting neural network [abstract]. Lin teaches wherein:
(a)
the input image comprises an obscured interaction region between the entity and an additional entity.
In further view of Lin, Lin teaches with reference to Fig. 2, the panoptic inpainting system (102) receives a request to generate an inpainted digital image. The panoptic inpainting system (102) identifies the designated area of missing or flawed pixels indicated by a binary mask obfuscating or occluding a portion of the digital image depicting four women (said obscured interaction region) against a mountain backdrop [0061].
It would have been obvious before the effective filing date of the claimed invention to of Men in view of Zhou to use the inpainting technology of Lin in order correct missing or flawed pixels, especially in the case of occluding areas of multiple objects, such as the four women [0061]. As Lin teaches, it is common in the cases where a designated area of a digital image depicts two objects that are adjacent to one another or that overlap by some portion, and further share a common semantic label, to be have the designated area be generated by one misshapen block of pixels that results from attempting to generate one object of the semantic label where the two separate objects should appear [Lin: 0036]. The method/system of Lin helps to maintain the boundary of the two objects depicted.
RE claim 6, Men in view of Zhou teaches a source parsing map is used to separate the full image into regions (said encoding first region of input image), so that Es independently encodes the styles in different semantics [Zhou: 0014]. However, Men in view of Zhou fail to provide further detail about multiple entities.
Lin is made of record as teaching a system/method for panoptically guiding digital image inpainting utilizing a panoptic inpainting neural network [abstract]. Lin teaches further comprising:
(a)
identifying a first bounding box for the entity and a second bounding box for an additional entity,
With reference to Fig. 2 of Lin, the panoptic inpainting system (102) receives a request to generate an inpainted digital image. The panoptic inpainting system (102) identifies the designated area of missing or flawed pixels indicated by a binary mask obfuscating or occluding a portion of the digital image depicting four women against a mountain backdrop [0061]. With reference to Fig. 4, the panoptic inpainting system (102) generates a crop of the predicted digital image (410) to focus on a particular object or region of pixels, e.g., the predicted crop (420) [0086]. The panoptic inpainting system (102) generates a binary mask to mask out the pixels around the individual depicted in the predicted crop (420) so that only the pixels representing the individual remain [0086]. The predicted crop area is a bounding box, as shown in Fig. 4 [0089]. Although the example of Fig. 4 illustrates a single entity, Lin further teaches determining multiple entities, such as shown in Fig. 7B, that determines the segmentation map [0112]. It is implied that the teachings of Fig. 4 would extend to input images
(b)
wherein the first region is based on the first bounding box.
In the combined invention, it would have been obvious before the effective filing date of the claimed invention to utilize the bounding boxes of Lin to determine the regions of Men in view of Zhou. As taught by Lin, the bounding boxes determine the entities based on the segmentation map similarly taught by Men in view of Zhou, i.e., source parsing map is used to separate the full image into regions, so that Es independently encodes the styles in different semantics [Zhou: 0014].
The same motivation to combine as taught in the rationales of claims 2 and 3 are incorporated herein.
Claim 4 is rejected under 35 U.S.C. 103 as being unpatentable over MEN et al. (“Controllable Person Image Synthesis with Attribute-Decomposed GAN”) in view of ZHOU et al. (“Cross Attention Based Style Distribution for Controllable Person Image Synthesis”) as applied to claim 1, and in further view of NINA PARAVECINO et al. (referenced as “Nina” throughout) (US 2021/0104029 A1).
RE claim 4, Men in view of Zhou teaches the limitations of claim 4 with the exception of discussing combining images. Nina is made of record as teaching a technique for synthesizing an image of a person in an unseen pose [abstract]. Fig. 1, apparatus (100) generates or attains input images (111) [0031]. Input images (111) are pictures of sequences of video pictures captured from different viewpoints [0031]. The techniques of Nina results in images of persons in scene (110) may be synthesized such that the persons have a pose matching a selected target pose (e.g., a user or application selected pose) to further enhance the immersive experience [0036]. Segmentation, pose estimation, and pose synthesis may be applied to any image or images of input images (111) [0032]. Segmentation module (102) generates a background/foreground (BG/FG) mask (112) using a CNN [0034]. BG/FG mask (112) detects e.g., players and other pertinent objects from a background [0035]. BG/FG mask (112) may indicate any number of foreground objects or persons [0037]. Pose estimation module (103), which generates pose data (113) for one or more persons represented by input image (111) [0038-0039]. Pose data (113) includes keypoint data including locations of such keypoints and their corresponding body parts [0039]. Nina further teaches provided target pose (114) for which it is desirable of a person of input image (111) to be synthesized [0043]. Pose synthesizer (104) generates a synthesized pose image (115) such that synthesized pose image (115) includes a representation of the person of interest in target pose (114) [0044]. Specifically, Nina teaches wherein obtaining the input image comprises:
(a)
obtaining a first preliminary image depicting the entity and a second preliminary image depicting an additional entity; and
A body part occlusion for a body part of the representation of the person is then detected and rectified by identifying a second image (said second preliminary image) corresponding to a second view of the scene having a second representation of the first body part of the person by identifying another image corresponding to another view of the scene having a second representation of the body part of the person [0044]. Thus, the initial input image (111) would correspond to the claimed first preliminary image [0045].
(b)
combining the first preliminary image and the second preliminary image to obtain the input image.
Nina further teaches projecting the image of the body part onto the portion of the image including the person [0044]. A geometric transformation is then applied to the merged image including the reprojected body part to generate a synthesized image including a representation of the person in target pose (114) such that the geometric transformation is based on the body part image and target pose (114) [0044].
It would have been obvious before the effective filing date of the claimed invention to provide the system of Nina with the combined image synthesis of Men in view of Zhou. As Nina teaches, it is advantageous to translate a target pose to a person that has not been imaged in the target pose. Such synthesized poses or synthesized images may be used in a variety of contexts [0025].
Claim 8 is rejected under 35 U.S.C. 103 as being unpatentable over MEN et al. (“Controllable Person Image Synthesis with Attribute-Decomposed GAN”) in view of ZHOU et al. (“Cross Attention Based Style Distribution for Controllable Person Image Synthesis”) as applied to claim 1, and in further view of ZHENG et al. (US 2024/0169500 A1).
RE claim 8, Men in view of Zhou teaches the limitations of claim 8 with the exception of discussing a noisy image. Zheng in made of record as teaching system/methods for image processing that results in a final output image that includes inpainted content [abstract]. Zheng teaches wherein generating the output image comprises:
(a)
obtaining a noisy image; and
Fig. 9, Zheng teaches the system receives an image including a first region that includes content and a second region to be inpainted [0098]. At operation (910), the system adds noise to the image to obtain a noisy image [0099]. Diffusion model (300) generates a noisy image by iteratively adding random noise to the masked image [0099].
(b)
performing a diffusion process on the noisy image to obtain the output image.
The system of Zheng generates a set of intermediate output images based on the noisy image using a diffusion model, where the diffusion model is trained using a perceptual loss, and where each of the set of intermediate output images includes an intermediate prediction of a final output image based on a corresponding intermediate noise level of the diffusion model [0101-0104]. The system generates the final output image based on the intermediate output image using the diffusion model, where the final output image includes inpainted content in the second region that is consistent with the content in the first region [0105].
It would have been obvious before the effective filing date of the claimed invention to implement the diffusion model of Zheng with the method/system of Men in view of Zhou because it allows for an image that includes a masked area to be modified to include inpainted content [Zheng: 0022]. This is especially helpful when objects overlap and portions are obscured by each other.
Claim 9 is rejected under 35 U.S.C. 103 as being unpatentable over MEN et al. (“Controllable Person Image Synthesis with Attribute-Decomposed GAN”) in view of ZHOU et al. (“Cross Attention Based Style Distribution for Controllable Person Image Synthesis”) as applied to claim 1, and in further view of YAMAZAKI (US 2022/0343667 A1).
RE claim 9, Men teaches a pretrained human parser [0006] along with training using a target pose (Pt) and a source person image (Is) [0012]. Zhou further teaches a single training stage [0004]. However, Men in view of Zhou fail to provide details on the training model. Yamazaki teaches a system/method for human detection in an image [abstract].
(a)
the image generation model is trained using a training set including a training image and a training skeleton map,
Fig. 4 of Yamazaki shows a skeleton detection unit (34) that can use a DNN having CNN architecture [0048-0049]. The skeleton detect is trained in advance with a large number of training images representing human skeletons. The skeleton detector outputs the positions of reference points for determining a skeleton [0049].
(b)
wherein the training image includes a plurality of entities and an obscured interaction region between the plurality of entities, and
Yamazaki further teaches a masking unit (35) that modifies the integrated human region (said obscured interaction region) by masking a region representing the newly detected foremost person in the integrated human region, based on the skeleton information [0051].
(c)
wherein the training skeleton map includes pose information for the plurality of entities.
Fig. 4 of Yamazaki shows a skeleton detection unit (34) that can use a DNN having CNN architecture [0048-0049]. The skeleton detect is trained in advance with a large number of training images representing human skeletons. The skeleton detector outputs the positions of reference points for determining a skeleton [0049].
It would have been obvious before the effective filing date of the claimed invention to utilize the training of Yamazaki within the system/method of Men in view of Zhou. As taught above, Men in view of Zhou mentions training images, but does not provide further detail. Yamazaki teaches further details on how training can occur. This provides a means for detecting a human within an image [Yamazaki: 0076].
Claims 15, 16, 20 are rejected under 35 U.S.C. 103 as being unpatentable over MEN et al. (“Controllable Person Image Synthesis with Attribute-Decomposed GAN”) in view of LIN et al. (US 2024/0127410 A1).
RE claim 15, Men teaches a method that allows for the synthesis of more realistic output images and automatic separation of un-annotated attributes [0001]. Men teaches an apparatus for image generation, comprising:
(i)
at least one processor; at least one memory component coupled with the at least one processor; and
Men illustrates in Fig. 2 an overview of network architecture of the generator proposed [0013]. It is implied that such a generator would include a processor and memory to carry out the function and computing as discussed.
(a)
and an image generation model comprising parameters stored in the at least one memory component and trained to receive an input image and pose information for a plurality of entities in the input image and
Fig. 2, source person image (Is) (said input image) [0013]. Men teaches the corresponding keypoint-based pose (P) (said pose information), 18 channel heat map that encodes the locations of 18 joints of a human body, can be automatically extracted via existing pose estimation method [0012]. Men fails to discuss the situation when there are multiple entities in the input image.
With reference to Fig. 2 of Lin, the panoptic inpainting system (102) receives a request to generate an inpainted digital image. The panoptic inpainting system (102) identifies the designated area of missing or flawed pixels indicated by a binary mask obfuscating or occluding a portion of the digital image depicting four women (said plurality of entities) against a mountain backdrop [0061].
(b)
to generate an output image depicting an interaction between the plurality of entities based on the pose information.
Fig. 2 of Men illustrates the generator whose inputs are the target pose (Pt) and a source person image (Is), and the output is the generated image (Ig) with source person (Is) in the target pose (Pt) [0013]. The desired person image is reconstructed from target features by a decoder (Ig) [0013]. In further view of Lin, Lin teaches the panoptic inpainting system (102) performs an act (208) to generate inpainted digital image [0064].
It would have been obvious before the effective filing date of the claimed invention of Men to use the inpainting technology of Lin in order correct missing or flawed pixels, especially in the situation of occluding areas of multiple objects, such as the four women [0061]. As Lin teaches, it is common in the cases where a designated area of a digital image depicts two objects that are adjacent to one another or that overlap by some portion, and further share a common semantic label, to be have the designated area be generated by one misshapen block of pixels that results from attempting to generate one object of the semantic label where the two separate objects should appear [Lin: 0036]. The method/system of Lin helps to maintain the boundary of the two objects depicted.
RE claim 16, in further view of Lin, Lin is made of record as teaching a system/method for panoptically guiding digital image inpainting utilizing a panoptic inpainting neural network [abstract]. Lin teaches wherein:
(a)
the image generation model comprises a boundary component configured to identify a bounding box for each of the entities.
With reference to Fig. 2 of Lin, the panoptic inpainting system (102) receives a request to generate an inpainted digital image. The panoptic inpainting system (102) identifies the designated area of missing or flawed pixels indicated by a binary mask obfuscating or occluding a portion of the digital image depicting four women against a mountain backdrop [0061]. With reference to Fig. 4, the panoptic inpainting system (102) generates a crop of the predicted digital image (410) to focus on a particular object or region of pixels, e.g., the predicted crop (420) [0086]. The panoptic inpainting system (102) generates a binary mask to mask out the pixels around the individual depicted in the predicted crop (420) so that only the pixels representing the individual remain [0086]. The predicted crop area is a bounding box, as shown in Fig. 4 [0089]. Although the example of Fig. 4 illustrates a single entity, Lin further teaches determining multiple entities, such as shown in Fig. 7B, that determines the segmentation map [0112]. It is implied that the teachings of Fig. 4 would extend to input images
The same motivation to combine as taught in the rationales of claims 2 and 3 are incorporated herein.
RE claim 20, Men teaches wherein:
(a)
the image generation model comprises a U-net architecture.
Fig. 2 of Men teaches a U-Net architecture [0013].
Claims 17, 18 are rejected under 35 U.S.C. 103 as being unpatentable over MEN et al. (“Controllable Person Image Synthesis with Attribute-Decomposed GAN”) in view of LIN et al. (US 2024/0127410 A1) as applied to claim 15, and in further view of ZHOU et al. (“Cross Attention Based Style Distribution for Controllable Person Image Synthesis”).
RE claim 17, Fig. 2 of Men illustrates the generator whose inputs are the target pose (Pt) and a source person image (Is), and the output is the generated image (Ig) with source person (Is) in the target pose (Pt) [0013]. The desired person image is reconstructed from target features by a decoder (Ig) [0013]. However, Men nor Lin discuss image features. Zhou teaches wherein:
(a)
the image generation model comprises a cross-attention layer configured to perform a cross-attention mechanism between image features of the input image and features representing the plurality of entities to obtain modified image features.
Zhou teaches a cross attention based style distribution (CASD) module that computes between the source semantic styles and target pose for pose transfer [0001, 0004, 0006]. The module intentionally selects the style represented by each semantic and distributes them according to the target pose [0001]. Zhou teaches given a source image (Is) (said input image) under the pose (Ps) the system/method of Zhou synthesizes a high fidelity image (Ît) (said modified image) under a different target pose (Pt) [0013]. As shown in Fig. 2, the desired pose (Pt) is directly used as the input by encoder (Ep), which describes the key point positions of human body (said entity features) [0014]. To facilitate accurate style extractions (said image features from input image) from source image (Is), the method/system of Zhou employs source parsing map (Ss) to separate the full image into regions, so that Es independently encodes the styles in different semantics [0014]. As shown in Fig. 2, Fp represents pose features from Ep (said entity features) and Fs represents style features from Es (said image features) [0014]. They are utilized by the CASD blocks [0014]. The cross attention module adapts the source style into the required pose (F’s) [Fig. 3, 0020].
It would have been obvious before the effective filing date of the claimed invention to utilize the CASD model of Zhou with the image synthesis of Men in view of Lin because the attention matrix in cross attention expresses the dynamic similarities between the target pose and the source styles for all semantics. Therefore, it can be utilized to route the color and texture from the source image, and is further constrained by the target parsing map to achieve a clearer objective [Zhou: 0001].
RE claim 18, in further view of Zhou, Zhou teaches wherein:
(a)
the cross-attention layer is configured to compute a key vector and a value vector for each of the plurality of entities.
Zhou teaches using keys and values within the self-attention modules [0012]. The computing is also described in [0018-0021].
The same motivation to combine as taught in the rationale of claim 17 is incorporated herein.
Claim 19 is rejected under 35 U.S.C. 103 as being unpatentable over MEN et al. (“Controllable Person Image Synthesis with Attribute-Decomposed GAN”) in view of LIN et al. (US 2024/0127410 A1) as applied to claim 15, and in further view of ZHENG et al. (US 2024/0169500 A1).
RE claim 19, Men in view of Lin teaches the limitations of claim 19 with the exception of discussing a diffusion model. Zheng in made of record as teaching system/methods for image processing that results in a final output image that includes inpainted content [abstract]. Zheng teaches wherein:
(a)
the image generation model comprises a diffusion model.
Fig. 9, Zheng teaches the system receives an image including a first region that includes content and a second region to be inpainted [0098]. At operation (910), the system adds noise to the image to obtain a noisy image [0099]. Diffusion model (300) generates a noisy image by iteratively adding random noise to the masked image [0099]. The system of Zheng generates a set of intermediate output images based on the noisy image using a diffusion model, where the diffusion model is trained using a perceptual loss, and where each of the set of intermediate output images includes an intermediate prediction of a final output image based on a corresponding intermediate noise level of the diffusion model [0101-0104]. The system generates the final output image based on the intermediate output image using the diffusion model, where the final output image includes inpainted content in the second region that is consistent with the content in the first region [0105].
It would have been obvious before the effective filing date of the claimed invention to implement the diffusion model of Zheng with the method/system of Men in view of Lin because it allows for an image that includes a masked area to be modified to include inpainted content [Zheng: 0022]. This is especially helpful when objects overlap and portions are obscured by each other.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MICHELLE L SAMS:
direct telephone number:
(571) 272-7661
email:
michelle.sams@uspto.gov
The examiner is currently part time and can be reached Mon.-Fri. 5:30am-9:30am.
Examiner interviews are available via telephone 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, Kee M. Tung can be reached on (571)272-7794. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/MICHELLE L SAMS/
Primary Examiner, Art Unit 2611
29 May 2026
Prosecution Insights