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 .
DETAILED ACTION
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.
Claims 1-2, 5, 7-8, and 11-20 are rejected under 35 U.S.C. 103 as being unpatentable over Cheng (Pub No. US 20250272887 A1) in view of Sunkavalli (Pub No. US 20200273237 A1).
As per claim 1, Cheng teaches the claimed:
1. A method for image generation, comprising:
obtaining an input image and an input prompt, wherein the input image depicts an object and the input prompt describes a lighting condition for the object; (Cheng [0057]: “According to an example implementation of the present disclosure, a prompt may be specified, for example, “a girl sitting on the hill looking at the sky, with her hair blowing in a clear day, with a cat . . . ”. Further, the prompt may specify generating a plurality of images with the following resolutions: 1536×1536, 384×384, 256×256, and 288×512. Images 620, 622, 624, and 626 represent images of respective resolutions generated using the proposed technical solutions, the content of these images is realistic and color bright, and the overall visual effect is greatly superior to images 610 through 616. It should be understood that while a prompt is provided in English as an example, a prompt may alternatively and/or additionally be written in other languages (e.g., Chinese, French, etc.).” The prompt results in a superior quality and mentions brightness. This is a lighting characteristic. Cheng figs. 2 and 3 show images being input into the model along with the prompt specifying a quality of the image before outputting a generated one. One of these is the input image.).
Cheng alone does not explicitly teach the remaining claim limitations.
However, Cheng in combination with Sunkavalli teaches the claimed:
generating, using a low-rank adaptation layer of an image generation model, relighted image features based on the input image and the input prompt, wherein the relighted image features represent the object with the lighting condition; (Cheng [0044]: “A fine-tuning plug-in (such as the LoRA plug-in 420 in FIG. 4A) may be implemented based on a low-rank adaptation of Large Language Models (LoRA) technical solution. LoRA may implement customization requirements (e.g., generate images of a specified style, etc.) with a small amount of data without modifying the backbone model parameters of the SD, the required training resource is much less than training SD model. The parameters of the LoRA may be injected into the SD model, changing a portion of the functionality of the SD model, e.g., generating an image with a specified style, and so on. According to an example implementation of the present disclosure, the LoRA plug-in 420 supporting multi-resolution adaptation may be injected into the first machine learning model 210, and then the first machine learning model is converted into a second machine learning model supporting multi-resolution image generation.” Cheng teaches a specified style for an image generation, as described above. This can include brightness. Cheng discusses generating images with certain qualities such as brightness in mind based on prompts. Cheng [0057]: “According to an example implementation of the present disclosure, a prompt may be specified, for example, “a girl sitting on the hill looking at the sky, with her hair blowing in a clear day, with a cat . . . ”. Further, the prompt may specify generating a plurality of images with the following resolutions: 1536×1536, 384×384, 256×256, and 288×512. Images 620, 622, 624, and 626 represent images of respective resolutions generated using the proposed technical solutions, the content of these images is realistic and color bright, and the overall visual effect is greatly superior to images 610 through 616. It should be understood that while a prompt is provided in English as an example, a prompt may alternatively and/or additionally be written in other languages (e.g., Chinese, French, etc.).” Additionally, Sunkavalli teaches relighted an image based on a description input into a model. Sunkavalli [0006]: “One or more embodiments described herein provide benefits and/or solve one or more of the foregoing or other problems in the art with systems, methods, and non-transitory computer readable storage media that train and utilize a deep-learning neural network model to generate digital images portraying objects illuminated under novel lighting based on a small sample of input digital images portraying the objects under calibrated lighting. For example, in one or more embodiments the disclosed systems utilize an object relighting neural network trained to generate a target digital image of an object illuminated from a target lighting direction based on five or fewer input digital images. To illustrate, in one or more embodiments, the disclosed systems train an object relighting neural network based on training digital images, training lighting directions, and ground truth digital images. Upon training, the disclosed systems can identify a set of input digital images portraying an object. The object relighting neural network then utilizes the trained object relighting neural network to generate a target digital image that portrays the object illuminated from a target lighting direction. In this manner, the disclosed systems can efficiently, accurately, and flexibly generate digital images illuminated under different lighting conditions with a sparse number of initial digital images, even for digital images that include complex geometric shapes, materials, and lighting effects.”).
and generating, using the image generation model, a synthetic image based on the relighted image features, wherein the synthetic image depicts the object with the lighting condition.
(Cheng [0004]: “In a first aspect of the present disclosure, a method for generating an image is provided. In the method, a first machine learning model is obtained, the first machine learning model being obtained based on a reference image having a first resolution. The first machine learning model is fine-tuned to a second machine learning model by a fine-tuning plug-in that is obtained based on a reference image having the second resolution. A target image is generated based on a target prompt by a second machine learning model, the target image having a resolution and image content specified by the target prompt.”. The target image is the synthetic image. The change in resolution quality can be combined with relighting using a learning model as taught by Sunkavalli above. Sunkavalli [0006]: “One or more embodiments described herein provide benefits and/or solve one or more of the foregoing or other problems in the art with systems, methods, and non-transitory computer readable storage media that train and utilize a deep-learning neural network model to generate digital images portraying objects illuminated under novel lighting based on a small sample of input digital images portraying the objects under calibrated lighting. For example, in one or more embodiments the disclosed systems utilize an object relighting neural network trained to generate a target digital image of an object illuminated from a target lighting direction based on five or fewer input digital images. To illustrate, in one or more embodiments, the disclosed systems train an object relighting neural network based on training digital images, training lighting directions, and ground truth digital images. Upon training, the disclosed systems can identify a set of input digital images portraying an object. The object relighting neural network then utilizes the trained object relighting neural network to generate a target digital image that portrays the object illuminated from a target lighting direction. In this manner, the disclosed systems can efficiently, accurately, and flexibly generate digital images illuminated under different lighting conditions with a sparse number of initial digital images, even for digital images that include complex geometric shapes, materials, and lighting effects.”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the relighting of an image based on input lighting characteristics as taught by Sunkavalli with the system of Cheng in order to allow the LoRA method of generating a modified image of Cheng to be used to relight an image based on a prompt indicating lighting characteristcs.
As per claims 12 and 18, these claims are similar in scope to limitations recited in claim 1, and thus are rejected under the same rationale. As per claim 12, Cheng teaches a non-transitory computer-readable medium. Cheng [0106]: “According to example implementations of the present disclosure, there is provided a computer-readable storage medium having computer-executable instructions stored thereon, wherein the computer-executable instructions are executed by a processor to implement the method described above. According to example implementations of the present disclosure, a computer program product is further provided, the computer program product being tangibly stored on a non-transitory computer-readable medium and including computer-executable instructions, the computer-executable instructions being executed by a processor to implement the method described above. According to example implementations of the present disclosure, there is provided a computer program product having stored thereon a computer program, which when executed by a processor, implements the method described above.”
As per claim 2, Cheng alone does not explicitly teach the claimed limitations.
However, Cheng in combination with Sunkavalli teaches the claimed:
2. The method of claim 1, wherein: the lighting condition includes at least one of a color, a brightness, a shadow, and a reflective property. (Sunkavalli [0031]: “The image relighting system provides several advantages over conventional systems. For example, the image relighting system improves accuracy of implementing computing systems. In particular, by training an object relighting neural network to generate target digital images, the image relighting system improves the accuracy of target digital images portraying objects illuminated from target lighting directions. For example, using a trained object relighting neural network allows the image relighting system to avoid assumptions about the properties of an object that may fail as those properties increase in complexity. Moreover, by utilizing a trained object relighting neural network, the image relighting system can accurately replicate sophisticated lighting effects, such as specularities, shadows, or reflections in relation to intricate or composite objects portrayed in digital images.”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the feature of brightness to change the image as taught by Sunkavalli with the system of Cheng in order to a LoRA model to output a modified image with the shadows changed.
As per claim 5, Cheng teaches the claimed:
5. The method of claim 1, wherein: the image generation model is trained based on a pre-trained image generation model by adding the low-rank adaptation layer to the pre-trained image generation model. (Cheng [0044]: “A fine-tuning plug-in (such as the LoRA plug-in 420 in FIG. 4A) may be implemented based on a low-rank adaptation of Large Language Models (LoRA) technical solution. LoRA may implement customization requirements (e.g., generate images of a specified style, etc.) with a small amount of data without modifying the backbone model parameters of the SD, the required training resource is much less than training SD model. The parameters of the LoRA may be injected into the SD model, changing a portion of the functionality of the SD model, e.g., generating an image with a specified style, and so on. According to an example implementation of the present disclosure, the LoRA plug-in 420 supporting multi-resolution adaptation may be injected into the first machine learning model 210, and then the first machine learning model is converted into a second machine learning model supporting multi-resolution image generation.” The first machine learning model is the pre-trained model. The LoRA plug-in is added to the model.).
As per claim 7, Cheng alone does not explicitly teach the claimed limitations.
However, Cheng in combination with Sunkavalli teaches the claimed:
7. The method of claim 1, wherein generating the synthetic image comprises: computing a color transformation function based on the relighted image features, wherein the synthetic image is based on the color transformation function. (Sunkavalli [0068]: “By combining the two direction channels with the three color channels, the image relighting system generates a five-channel input for each input digital image. Thus, the image relighting system provides a 5k-channel input to the object relighting neural network. In some embodiments, the image relighting system uses input digital images that include fixed lighting directions (i.e., all sets of input digital images are captured using the same four or five lighting directions); therefore, the image relighting system configures the object relighting neural network to inherently process the input digital images as if they portrayed the fixed lighting directions. In such embodiments, the image relighting system does not provide the lighting directions for the input digital images. In such embodiments, the image relighting system provides a 3k-channel input that includes the color values of the input digital images.” The combination of the color channels and the lighting directions transforms the perceived colors of the scene. This is the color transformation function.).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the model for changing color based on lighting as taught by Sunkavalli with the system of Cheng in order to clearly convey the color output to the final generated image.
As per claim 8, Cheng alone does not explicitly teach the claimed limitations.
However, Cheng in combination with Sunkavalli teaches the claimed:
8. The method of claim 7, further comprising: predicting one or more color parameters based on the color transformation function, wherein the synthetic image is based on the one or more color parameters. (Sunkavalli [0025]: “In one or more embodiments, the image relighting system provides the set of input digital images to the object relighting neural network by generating and providing sets of color channels corresponding to the set of input digital images. For example, the image relighting system can generate and provide a first set of color channels corresponding to a first input digital image and a second set of color channels corresponding to a second input digital image. In particular, each set of color channels comprises color values reflecting pixels of the respective input digital image. In one or more embodiments, each set of color channels comprises three color channels, each channel including color values for a color within the RGB color model.” The color channels are the color parameters.).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the color parameters to determine the color of the newly generated image as taught by Sunkavalli with the system of Cheng in order to have a rigorous system for calculating the colors in the output image based on the input parameters of the model.
As per claims 13 and 20, this claim is similar in scope to limitations recited in a combination of claims 7 and 8, and thus are rejected under the same rationale.
As per claim 11, Cheng teaches the claimed:
11. The method of claim 1, wherein generating the synthetic image comprises: obtaining a noise map; (Cheng teaches a feature map. Cheng [0064]: “Details regarding the various steps of image generation have been described above, and in the following, an overall process for generating an image using a machine learning model will be described. The text-to-image model and corresponding personalization techniques may support generating high quality, imaginative images. However, since the receptive field of the convolutional layer in the U network of the diffusion model does not match the feature map size of the image, and the normalization cannot accommodate the statistical distribution of the feature maps in images with a plurality of resolutions, the quality will be significantly reduced when the resolution of the image generated by the model is away from the resolution of the training image.” Cheng teaches identifying noise. Cheng [0070]-[0071]: “[0070] The training objective in this formulation is minimizing the squared error between the gaussian noise and the estimated noise of the noise-added samples.
[0071] To reduce the training cost of the diffusion model and generate a high-resolution image, the SD model may encode the image using a variational autoencoder. SD performs forward and reverse denoising in latent space. Specifically, given the data x.sub.0˜q.sub.data(x), the encoder ε encodes the image into z.sub.0=ε(x.sub.0). For latent representations generated by the diffusion model in latent space, the decoder [AltContent: rect] may reconstruct it as an image. In SD, the encoder typically down-samples an image by a factor f=8. A loss function may be obtained.”
It would be obvious that the noise is present in the feature map, and could be used to create a noise map.).
and denoising the noise map based on the input prompt to obtain the synthetic image. (Cheng [0068]: “The generation process of the diffusion model includes forward diffusion and reverse denoising processes. Given one data sample X.sub.0˜q.sub.data(x), the diffusion model progressively injects small Gaussian noise into the data and generates samples by inverse denoising. Specifically, the forward diffusion process of the diffusion model is controlled by the Markov chain as q(x.sub.t|x.sub.t-1)=[AltContent: rect](x.sub.t; √{square root over (1−β)}.sub.tx.sub.t-1, β.sub.tI), wherein Pt is a variance schedule between 0 and 1. By using the reparameterization technique, the data distribution q.sub.data(x) may be converted to a marginal distribution q(x.sub.t|x.sub.0). q(x.sub.t|x.sub.0)=[AltContent: rect](x.sub.t; √{square root over (α.sub.t)}x.sub.0, (1−α.sub.t)I) may be obtained by using symbols: α.sub.t: =1−β.sub.t, and α.sub.t: =Π.sub.s=1.sup.tα.sub.s.”).
As per claim 16, Cheng teaches the claimed:
16. The non-transitory computer readable medium of claim 15, wherein: the image generation model is trained based on a reconstruction loss. (Cheng [0071]: “To reduce the training cost of the diffusion model and generate a high-resolution image, the SD model may encode the image using a variational autoencoder. SD performs forward and reverse denoising in latent space. Specifically, given the data x.sub.0˜q.sub.data(x), the encoder ε encodes the image into z.sub.0=ε(x.sub.0). For latent representations generated by the diffusion model in latent space, the decoder may reconstruct it as an image. In SD, the encoder typically down-samples an image by a factor f=8. A loss function may be obtained:”).
As per claim 19, Cheng alone does not explicitly teach the claimed limitations.
However, Cheng in combination with Sunkavalli teaches the claimed:
19. The system of claim 18, wherein: the low-rank adaptation layer comprises image relighting parameters stored in the memory component. (Sunkavalli [0062]: “The image relighting system can train the object relighting neural network 306 based on the determined loss. For example, in one or more embodiments, the image relighting system back propagates the determined loss to the object relighting neural network 306 to modify its parameters. In one or more embodiments, the image relighting system modifies the parameters of each layer of the object relighting neural network. Consequently, with each iteration of training, the image relighting system gradually increases the accuracy of the object relighting neural network 306 (e.g., through gradient assent or gradient descent). As shown, the image relighting system can thus generate the trained object relighting neural network 314.”
The image generation network that stores the parameters can be combined with the LoRA of Cheng. Cheng [0044]: “A fine-tuning plug-in (such as the LoRA plug-in 420 in FIG. 4A) may be implemented based on a low-rank adaptation of Large Language Models (LoRA) technical solution. LoRA may implement customization requirements (e.g., generate images of a specified style, etc.) with a small amount of data without modifying the backbone model parameters of the SD, the required training resource is much less than training SD model. The parameters of the LoRA may be injected into the SD model, changing a portion of the functionality of the SD model, e.g., generating an image with a specified style, and so on. According to an example implementation of the present disclosure, the LoRA plug-in 420 supporting multi-resolution adaptation may be injected into the first machine learning model 210, and then the first machine learning model is converted into a second machine learning model supporting multi-resolution image generation.”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the storing of relighting parameters in memory for use by a machine learning model as taught by Sunkavalli with the system of Cheng in order to allow storage of those parameters for use by the LoRA of Cheng.
Claims 3-4, and 6 is rejected under 35 U.S.C. 103 as being unpatentable over Cheng in view of Sunkavalli and further in view of Saharia (Pub No. US 20230377226 A1).
As per claim 3, Cheng alone does not explicitly teach the claimed limitations.
However, Cheng in combination with Saharia teaches the claimed:
3. The method of claim 1, wherein generating the relighted image features comprises: encoding the input prompt to obtain a prompt embedding, wherein the relighted image features are based on the prompt embedding. (Saharia abstract: “Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating images. In one aspect, a method includes: receiving an input text prompt including a sequence of text tokens in a natural language; processing the input text prompt using a text encoder neural network to generate a set of contextual embeddings of the input text prompt; and processing the contextual embeddings through a sequence of generative neural networks to generate a final output image that depicts a scene that is described by the input text prompt..” This can be combined with the generation of a relighted image as taught by Sunkavalli with the model of Cheng.).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the embedding of a prompt in vector form as taught by Saharia with the system of Cheng and Sunkavallli in order to convert the prompt into a mathematical and machine-readable representation for processing by the model.
As per claim 4, Cheng teaches the claimed:
4. The method of claim 3, wherein: the input prompt comprises a text prompt or an image prompt. (Cheng teaches receiving a text prompt. Cheng [0057]: “According to an example implementation of the present disclosure, a prompt may be specified, for example, “a girl sitting on the hill looking at the sky, with her hair blowing in a clear day, with a cat . . . ”. Further, the prompt may specify generating a plurality of images with the following resolutions: 1536×1536, 384×384, 256×256, and 288×512. Images 620, 622, 624, and 626 represent images of respective resolutions generated using the proposed technical solutions, the content of these images is realistic and color bright, and the overall visual effect is greatly superior to images 610 through 616. It should be understood that while a prompt is provided in English as an example, a prompt may alternatively and/or additionally be written in other languages (e.g., Chinese, French, etc.).”).
As per claim 6, Cheng alone does not explicitly teach the claimed limitations.
However, Cheng in combination with Saharia teaches the claimed:
6. The method of claim 1, wherein generating the relighted image features comprises: encoding the input image to obtain an image embedding, wherein the relighted image features are based on the image embedding. ((Saharia abstract: “Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating images. In one aspect, a method includes: receiving an input text prompt including a sequence of text tokens in a natural language; processing the input text prompt using a text encoder neural network to generate a set of contextual embeddings of the input text prompt; and processing the contextual embeddings through a sequence of generative neural networks to generate a final output image that depicts a scene that is described by the input text prompt.” Saharia teaches generating images to change lighting features. Saharia [0068]: “In some implementations, the output image 106 is further processed by the post-processor 130 to generate a final image (x) 108. For example, the post-processor 130 can perform transformations on the output image 106 such as image enhancement, motion blur, filtering, luminance, lens flare, brightening, sharpening, contrast, among other image effects. Some or all of the transformations performed by the post-processor 130 may also be performed by the sequence 121 when the GNNs 120 are suitably trained (e.g., by a training engine). For example, the GNNs 120 can learn these transformations and associate them with respective text modifiers included in text prompts 102. In some implementations, the system 100 does not include the post-processor 130 and the output image 106 generated by the sequence 121 is the final image 108. Alternatively, system 100 can disable the post-processor 130 such that transformations performed on the output image 106 by the post-processor 130 are equivalent to the identity operation.” This can be combined with the generation of a relighted image as taught by Sunkavalli with the model of Cheng.).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the embedding of an image as a prompt to generate an image with different lighting features as taught by Saharia with the system of Cheng and Sunkavalli in order to allow an image to be input into the model in a format the model could read and the image indicate the characteristics of the model to be generated.
Claim 9 is rejected under 35 U.S.C. 103 as being unpatentable over Cheng in view of Sunkavalli and further in view of Smetanin (Pub No. US 20240412433 A1).
As per claim 9, Cheng alone does not explicitly teach the claimed limitations.
However, Cheng in combination with Smetanin teaches the claimed:
9. The method of claim 1, wherein generating the synthetic image comprises: generating a background of the synthetic image, wherein content of the background is described by the input prompt. (Smetanin describes a variety of text-based prompts for specific image characteristics, including background. Smetanin [0081]-[0091]: “[0081] With some embodiments, the image processing component may include a prompt constructor that generates a text-based prompt customized to the input image to provide as additional input to the machine learning model. Rather than selecting from predefined text prompts, the prompt constructor analyzes the input image to generate a descriptive prompt that characterizes the image content and attributes. This customized prompt allows guiding the transition in a way that better preserves important image features compared to a fixed template prompt. For example, the generated prompt may describe the number of people present, their poses and expressions, the lighting conditions, background scene, and other visual details. The machine learning model is trained to leverage these image descriptions to produce a more context-aware stylized output that transfers finer aspects of the original image. At least one part of the prompt includes an instruction relevant to the desired stylization, for example: [0082] “Apply the style of Monet to the image.” [0083] “Transform the image into a Van Gogh-like painting.” [0084] “Stylize the image using the brushwork of Picasso.” [0085] “Create a watercolor version of the image.” [0086] “Apply an impressionistic style to the image, inspired by Renoir.” [0087] “Make the image look like a charcoal sketch.” [0088] “Give the image a vintage, sepia-tone effect.” [0089] “Apply a comic book-style filter to the image.” [0090] “Add a pointillism effect to the image, similar to Seurat.” [0091] “Create a pop art-inspired version of the image, reminiscent of Warhol.” The generated prompt could be input into the model of Cheng.).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the background description input as taught by Smetenin with the system of Cheng in order to use a LoRA model to output images with the background customized.
Claim 10 is rejected under 35 U.S.C. 103 as being unpatentable over Cheng in view of Sunkavalli and further in view of Smetanin and further in view of Smetanin (Pub No. US 20240296606 A1) hereinafter named Smetanin ‘606.
As per claim 10, Cheng alone does not explicitly teach the claimed limitations.
However, Cheng in combination with Smetanin ‘606 teaches the claimed:
10. The method of claim 9, further comprising: generating the background of the synthetic image based on an editing mask. (Smetanin ‘606 teaches image generation based on input prompt text and images. Smetanin ‘606 [0131]: “In FIG. 10, it is shown that the automated image generator 404 is used to generate multiple variations, or options, based on the same text prompt 802, and these are presented to the user in the image selection interface 1000 as the candidate images 1002. Given that many generative AI tools are probabilistic in nature, they may not produce an exact output for a given input, but rather generate a distribution of possible outputs. In the case of automated image generation, a model may generate multiple images that are all plausible interpretations of the given prompt (as determined by the mode), e.g., with some variation in colors, textures, lighting, or other visual elements. This can provide a user with a useful technological tool to enable selection from several outputs and use of the selected output in a creative manner.”
Smetanain ‘606 teaches segmenting an image with a mask to indicate background area. Smetanin ‘606 [0177]: “To apply this augmentation, a segmentation mask may be generated, and the automatically generated image 2102 may be applied to the media content item based on the generated segmentation mask. For example, the image processing system 202 may use a segmentation mask to separate the foreground image area 1812 from the background image area 1814. This mask may comprise a binary image that identifies which pixels belong to the foreground and which belong to the background. Based on the mask, the background image area 1814 may then automatically be replaced with the automatically generated image 2102.”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the segmentation mask to show background area as taught by Smetanin ‘606 with the system of Cheng in order to determine which portion of an input image is the background and replace it with other backgrounds if the user desires to alter the background.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to THOMAS JOHN FOSTER whose telephone number is (571)272-5053. The examiner can normally be reached Mon, Fri 8:30-6. Tues-Thurs 7:30-5.
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, Daniel Hajnik can be reached at 571-272-7642. 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.
/THOMAS JOHN FOSTER/Examiner, Art Unit 2616
/HAI TAO SUN/Primary Examiner, Art Unit 2616
Prosecution Insights