CTNF 18/816,846 CTNF 101491 DETAILED ACTION Notice of Pre-AIA or AIA Status 07-03-aia AIA 15-10-aia The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA. Priority The present application claims benefit of provisional applications 63/539,936 filed on 09/22/2023, 63/539,949 filed on 09/22/2023, 63/545,249 filed on 10/23/2023, and 63/545,252 filed on 10/23/2023. Information Disclosure Statement The information disclosure statement(s) (IDS) submitted on 8/27/2024, 3/24/2025, 11/21/2025 are in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statements are being considered by the examiner. Claim Rejections - 35 USC § 112 07-30-02 AIA The following is a quotation of 35 U.S.C. 112(b): (b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention. The following is a quotation of 35 U.S.C. 112 (pre-AIA), second paragraph: The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention. 07-34-01 Claims 2, 8, 12, and 17 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention. Claims 2, 8, 12, and 17 recite the phrase “ luminance-chrominance (YUV) ”. Luminance-chrominance has other types of color spaces than solely YUV (e.g. YCbCr, YPbPr,YCoCg etc.). Including “YUV” in parenthesis behind “luminance-chromance” makes the limitations unclear on what is being claimed. It is unclear if the claims are referring to any type of luminance-chrominance color spaces or only YUV; therefore, the claims are indefinite. Claims 2, 8, 12, and 17 are rejected under 112(b) for being indefinite. Applicant may amend claims to read only “luminance-chrominance” or “YUV” and make appropriate corrections to related references to “luminance-chrominance” or “YUV” that may or may not be found throughout the remainder of the claims in order to overcome the rejection. Claim Rejections - 35 USC § 103 07-06 AIA 15-10-15 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. 07-20-aia AIA 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. 07-23-aia AIA The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows: 1. Determining the scope and contents of the prior art. 2. Ascertaining the differences between the prior art and the claims at issue. 3. Resolving the level of ordinary skill in the pertinent art. 4. Considering objective evidence present in the application indicating obviousness or nonobviousness. Claim(s) 1-5 and 11-15 are rejected under 35 U.S.C. 103 as being obvious over Besenbruch et al. (US 20230154055 A1; provided by applicant in IDS) in view of Yoshimura et al. ("DynamicISP: Dynamically Controlled Image Signal Processor for Image Recognition"; copy provided by Examiner). Regarding claims 1 (method) and claim 11 (electronic device), Besenbruch teaches: A method (Abstract “a computer-implemented method…”; And “An electronic device comprising: at least one processing device configured to:” (claim 11) perform the mirrored method limitations found below (Basenbruch [0074] “The method may be one wherein the user device is a laptop computer”; [0139] “the computer program product executable on a processor”)) comprising: obtaining a raw image (Abstract and [0007] (i) “receiving an input image”) ; mapping, using a raw image encoder, the raw image to a compressed domain, wherein the raw image is represented using latent variables in the compressed domain (Besenbruch teaches “(ii) encoding the input image using a first trained neural network, using the first computer system, to produce a latent representation; (iii) quantizing the latent representation using the first computer system to produce a quantized latent; (iv) entropy encoding the quantized latent into a bitstream, using the first computer system” ([0007]) and “The latent variables y are the quantized (integer rounded) outputs of a Encoder neural network” ([0775]). For sake of clarity, the “encoding the input image” taught by Besenbruch equates to mapping the image, the “first trained neural network/Encoder neural network” equates to the image encoder being used, the “latent representation” and “quantized latent” equates to the compressed domain because they are encoded representations produced by the compression architecture and are subsequently entropy coded into compressed bitstream, and as detailed in [0775], the latent variables are being used as output (i.e. representation). ). performing one or more image signal processing operations on the latent variables, wherein each of the one or more image signal processing operations is configured to operate in the compressed domain, and wherein the one or more image signal processing operations generate processed latent variables; and mapping, using an output image decoder, the processed latent variables to an output image in an output color space (Besenbruch teaches entropy decoding a bitstream to produce a quantized latent (i.e. latent variable) and using a second trained neural network to produce an output image from the quantized latent ([0007] “entropy decoding the bitstream to produce the quantized latent…using a second trained neural network to produce an output image from the quantized latent, wherein the output image is an approximation of the input image”), and [0075] (quoted in earlier limitation above) further teaches that latent variables y are the quantized outputs of the encoder neural network. Besenbruch further teaches image representations in the YCbCr color space and processing image information according to Y, Cb, and Cr color space components ([0583]; FIG. 13). Accordingly, Besenbruch teaches using an output image decoder neural network to map a latent variable representation (i.e. the quantized latent) to an output image in an output color space.) . Besenbruch fails to explicitly disclose: performing the tasks discussed above using an obtained raw input image, and while Besenbruch does teach latent variables in a compressed domain (See Besenbruch [0007] and [0075], discussed above.) , Besenbruch fails to explicitly disclose: performing one or more image signal processing operations on the latent variables, wherein each of the one or more image signal processing operations is configured to operate in the compressed domain, and wherein the one or more image signal processing operations generate processed latent variables. In a related art, Yoshimura teaches: obtaining and processing raw images (p. 1, left column, second sentence of Section 1 “converts raw out puts of the sensors, RAW images, into commonly used standard RGB (sRGB) images”) and performing image signal processing (ISP) using latent variables and generating updated latent variables through the image signal processing operations (p. 2 left column, lines 30-36, “A latent update style ISP controller that manages multiple ISP functions. It generates shared latent variables which indicate how to change the RAW images with the entire ISP pipeline; then, after applying each ISP function, sequentially updates them to contain information on what the remaining functions should do considering upstream functions.”). Thus, Yoshimura teaches performing image signal processing on latent variables and generating updates (i.e. processed) latent variables through the image signal processing operations. It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to apply the raw-image image signal processing (ISP) techniques of Yoshimura to the latent compressed domain representation of Besenbruch in order to improve the quality of the reconstructed output image by performing image enhancement operations on image information prior to image reconstruction. Applying Yoshimura’s ISP techniques to Besenbruch’s latent representation would predictably improve image reconstruction quality by enhancing image representation prior to image reconstruction, while maintaining the image information and quality in the compressed domain. Both references share the common objective of improving image quality using learned image representation; therefore, a person of ordinary skill in the art would have found Yoshimura’s image enhancement techniques applicable to Besenbruch’s latent image representations. Also, applying Yoshimura’s ISP techniques to Besenbruch’s model would further address current deficiencies in compressing image and video content over communications networks and the need to “increase the compression, while preserving displayed image quality, or to increase the displayed image quality, while not increasing the amount of data that is actually transmitted across the communications networks” (see Besenbruch [0004]). Regarding claims 2 and 12, Besenbruch and Yoshimura teach the method of claim 1, and the electronic device of claim 11. Yoshimura further teaches: wherein one of: the output color space is a red-green-blue (RGB) color space and the output image is an RGB image ; or the output color space is a luminance-chrominance (YUV) color space and the output image is a YUV image (Yoshimura teaches generating an output image in an RGB color space, teaching that the disclosed image signal processor “converts raw out puts of the sensors, RAW images, into commonly used standard RGB (sRGB) images” (p. 1, left column, second sentence of Section 1). A standard RGB (sRGB) is known in the art as both a specific type of RGB color space and an RGB image. Accordingly, Yoshimura teaches wherein the output color space is a RGB color space and the output image is an RGB image. Besenbruch and Yoshimura fail to disclose “the output color space is a luminance-chrominance (YUV) color space and the output image is a YUV image”. However, the claim, as currently written, only requires the output color space to be either a red-green-blue (RGB) color space and the output image to be a RGB image OR the output color space to be a luminance-chrominance (YUV) color space and the output image to be a YUV image.). Regarding claims 3 and 13, Besenbruch and Yoshimura teach the method of claim 1, and the electronic device of claim 11. Yoshimura further teaches: wherein the raw image comprises a Tetra image, a Bayer image, a hexa-deca image, or a red-green-blue-white (RGBW) image (Yoshima teaches using raw image data (p. 1, left column, second sentence of Section 1 “converts raw outputs of the sensors, RAW images, into commonly used standard RGB (sRGB) images”) and further provides an exemplary application using test images “taken with a RAW Bayer sensor” (p. 5, left column, line 6). Regarding claims 4 and 14, Besenbruch and Yoshimura teach the method of claim 1, and the electronic device of claim 11. Besenbruch further teaches: wherein at least one of: the raw image encoder comprises a convolutional neural network-based encoder with quantization; or the output image decoder comprises a convolutional neural network (Besenbruch teaches using “AI-based compression encoder-decoder pipelines,” including a first trained neural network encoder producing a latent representation and quantizing the latent representation to produce a quantized latent, and using a second neural network as a decoder to decode the bitstream to produce the quantized latent and produce an output image from the quantized latent (see [0007]; [0254]). Besenbruch further teaches the second trained neural network for producing an output image from the quantized latent is a convolutional network (see [0350]; claim 4). Regarding claims 5 and 15, Besenbruch and Yoshimura teach the method of claim 1, and the electronic device of claim 11, including “latent variables in the compressed domain”. Yoshimura further teaches: image signal processing functions including denoising and sharpening (see p. 5, left column, subsection “4.2 ISP functions” reads “We implement five ISP functions, auto gain (AG), denoiser (DN), sharpener (SN), gamma tone mapping (GM), and contrast stretcher (CS), in a differentiable manner.”) using latent variables that are updated throughout the ISP pipeline (p. 2 left column, lines 30-36) . Besenbruch and Yoshimura fail to explicitly teach: wherein the one or more image signal processing operations comprise at least one of: compressed domain denoising, compressed domain sharpening, compressed domain point spread function (PSF) inversion, compressed domain segmentation, compressed domain motion map estimation and motion compensation, or compressed domain motion image registration. However, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to apply the image signaling operations of denoising and/or sharpening taught by Yoshimura to the compressed domain latent representation previously taught by Besenbruch and Yoshimura in order to improve image quality of the reconstructed image by performing image signal processing on the image information represented in the latent compressed domain representation prior to reconstruction into a color space, which aligns with Besenbruch’s goal of reducing output image distortion and reducing file size (see Besenbruch [0008]) . Claim(s) 6 and 16 are rejected under 35 U.S.C. 103 as being obvious over Besenbruch et al. (US 20230154055 A1; provided by applicant in IDS) in view of Yoshimura et al. ("DynamicISP: Dynamically Controlled Image Signal Processor for Image Recognition"; copy provided by Examiner), in further view of Kang et al. (KR 2022008135 A; provided by applicant in IDS; see alternative translated copy provided by Examiner that is relied upon for cited quotations found below), in further view of Saripalli (US 20200272905 A1), and in further view of Dutta (US 20210265018 A1). Regarding claims 6 and 16, Besenbruch and Yoshimura teach the method of claim 1, and the electronic device of claim 11, including mapping a raw image to a compressed domain using a raw image encoder. Besenbruch and Yoshimura fail to explicitly disclose: wherein the raw image encoder comprises a trained machine learning model subjected to a training scheme comprising: training a teacher image encoder to map input images into the compressed domain, wherein the training yields trained network weights for the teacher image encoder; initializing network weights of the raw image encoder from the trained network weights for the teacher image encoder; and constraining outputs of the raw image encoder to be close to outputs of the teacher image encoder via a mean squared error loss in the compressed domain. In a related art, Kang teaches: a teacher-student training framework wherein a trained teacher network is to train a student network (see Kang Abstract). Specifically, Kang teaches “training a teacher model…and initially training a student model” ([0010]), wherein “Using a knowledge distillation technique, the weight parameters of the student network (200) can be learned based on the information of the teacher network (100) that has already been learned” ([0029]) . Kang further teaches retraining the student model using outputs of the teacher model by “modify the loss function of the student model using the output values of each residual block of the teacher model” ([0012]) and calculating “loss functions calculated based on the difference between the output feature maps of each filter of the student network (200) and the teacher network (100)” ([0068]). It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to apply the teacher training methodology of Kang in the raw image encoder of Besenbruch and Yoshimura (previously taught in claim 1) such that a raw image encoder is trained using information learned by a trained teacher encoder and further constrained to generate latent representations (i.e. “map input images,” as previously taught by Besenbruch in claim 1’s 103 rejection above, see Besenbruch [0007] and [0075] ) similar or close to those generated by the teacher encoder. A person of ordinary skill in the art would have recognized that applying the teacher-student training techniques of Kang to the latent compressed domain image representation and raw image encoder of Besenbruch and Yoshimura would predictably transfer learned compressed domain image representations from the teacher encoder to the raw image encoder, improving training and enable the raw image encoder to more accurately reproduce the teacher encoder’s compressed domain latent representation, and that such a model constitutes the composition of a training scheme. Besenbruch, Yoshimura, and Kang fail to explicitly disclose: initializing network weights of the raw image encoder from the trained network weights for the teacher image encoder and using a mean square loss function. In a related art, Saripalli teaches: initializing weights of an image encoder from a trained network weights of a teacher image encoder ([0041] “D3MC architecture 302 can then use the original weights/parameters of the teacher network 110 to initialize the student network 114.”). It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to initialize the raw image encoder using the trained weights of the teacher image encoder taught by Besenbruch, Yoshimura, and Kang using “the original weights/parameters of the teacher network…to initialize the student network” (Saripalli [0041]), thereby transferring previously learned representations from the trainer network to the student network and reducing training time. In a related art, Dutta teaches: another teacher-student model and further teaches a convolutional neural network training may utilize “mean-squared error loss” (see Dutta Abstract and [0418]-[0419]). It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to implement the teacher-student teacher feature map matching of Besenbruch, Yoshimura, Kang, and Saripalli using a mean squared error loss, as taught by Dutta, because mean squared error loss was a known loss function for minimizing differences between outputs of neural network models. All references utilize neural network training techniques involving image processing, learned feature representations, latent representations, or teacher-student models. Claim(s) 7 and 17 are rejected under 35 U.S.C. 103 as being obvious over Besenbruch et al. (US 20230154055 A1; provided by applicant in IDS) in view of Yoshimura et al. ("DynamicISP: Dynamically Controlled Image Signal Processor for Image Recognition"; copy provided by Examiner), in further view of Kang et al. (KR 2022008135 A; provided by applicant in IDS; see alternative translated copy provided by Examiner that is relied upon for cited quotations found below), in further view of Saripalli (US 20200272905 A1), in further view of Dutta (US 20210265018 A1), and in further view of “JPEG AI Common Training and Test Conditions” from ISO/IEC JTC 1/SC29/WG1 N100106 (provided by applicant in IDS; referred to below as “WG1”). Regarding claims 7 and 17, Besenbruch, Yoshimura, Kang, Saripalli, and Dutta teach the method of claim 6, and the electronic device of claim 17. While Besenbruch, Yoshimura, Kang, Saripalli, and Dutta teach (in claims 1 and 6s’ 103 rejections found above): a RGB color space and an output RGB image (see claim 1 above and Yoshimura p. 1, left column, second sentence of Section 1) , a teacher image encoder (see claim 6 above) and a training scheme comprising training a teacher image encoder to map input images into the compressed domain (see claim 6 above) , Besenbruch, Yoshimura, Kang, Saripalli, and Dutta fail to explicitly disclose: wherein: the teacher image encoder is a red-green-blue (RGB) image encoder; and the training scheme comprises training the teacher image encoder to map RGB images into the compressed domain. In a related art, WG1 (“JPEG AI Common Training and Test Conditions” reference) teaches: an image encoder is a red-green-blue (RGB) image encoder (WG1 teaches encoder receives RGB input images for subsequent processing (p. 7, section 2 “Format- PNG images (RGB color components, non-interlaced”; p. 8, first paragraph, “…the input of the encoder…must be in the PNG (RGB color space) format…the RGB decoded images will be used for subjective quality evaluation.”).) ; the image encoder maps RGB images into a compressed domain (WGI states “The scope of the JPEG AI is the creation of a learning-based image coding standard offering a single-stream, compact compressed domain” (p. 1, paragraph [0001]). Thereby, the image encoder receives the RGB images (as described above) and maps the RGB images into a compressed domain.) . It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings and training scheme model of Besenbruch, Yoshimura, Kang, Saripalli, and Dutta to incorporate the RGB image encoder of WG1 in order to more accurately map RGB images into a compressed domain. Doing so would preserve image quality (e.g. RGB quality) while decreasing bandwidth and storage requirements. All references utilize neural network training techniques involving image processing, learned feature representations, latent representations, or teacher-student models. Examiner acknowledges claim 17 claims: the teacher image encoder is a red-green-blue (RGB) image encoder or a luminance-chrominance (YUV) image encoder; and the training scheme comprises training the teacher image encoder to map RGB images or YUV images into the compressed domain. Therefore, the teachings of Besenbruch, Yoshimura, Kang, Saripalli, and Dutta (seen above) do read on the image encoder being a red-green-blue (RGB) encoder, and the training scheme comprising training the teacher image encoder to map RGB images. As written, claim 17 doesn’t require the luminance-chromance (YUV) image encoder or the mapping of YUV images into the compressed domain. Claim 8’s rejection below does however show support of prior art teachings of the image encoder being a luminance-chrominance (YUV) image encoder and the training scheme comprising training teacher encoder to map YUV images into the compressed domain. Claim 8 is rejected under 35 U.S.C. 103 as being obvious over Besenbruch et al. (US 20230154055 A1; provided by applicant in IDS) in view of Yoshimura et al. ("DynamicISP: Dynamically Controlled Image Signal Processor for Image Recognition"; copy provided by Examiner), in further view of Kang et al. (KR 2022008135 A; provided by applicant in IDS; see alternative translated copy provided by Examiner that is relied upon for cited quotations found below), in further view of Saripalli (US 20200272905 A1), in further view of Dutta (US 20210265018 A1), in further view of “JPEG AI Common Training and Test Conditions” from ISO/IEC JTC 1/SC29/WG1 N100106 (provided by applicant in IDS; referred to below as “WG1”), and in further view of Mancuso et al. (“A novel high-quality YUV-based image coding technique for efficient image storage in portable electronic appliances”; copy provided by Examiner). Regarding claim 8, Besenbruch, Yoshimura, Kang, Saripalli, and Dutta teach the method of Claim 6. While Besenbruch, Yoshimura, Kang, Saripalli, and Dutta teach (in claims 1 and 6s’ 103 rejections found above): a teacher image encoder (see claim 6 above) and a training scheme comprising training a teacher image encoder to map input images into the compressed domain (see claim 6 above) , Besenbruch, Yoshimura, Kang, Saripalli, and Dutta fail to explicitly disclose: wherein: the teacher image encoder is a luminance-chrominance (YUV) image encoder; and the training scheme comprises training the teacher image encoder to map YUV images into the compressed domain. In a related art, Mancuso teaches: teaches converting image data into YUV image representation (Mancuso, Abstract “We propose a YUV-based encoding strategy…”; p. 697, right column, second paragraph, “Every 24-bit RGB pixel is converted into a 24-bit YUV triplet”) and performing compression on the YUV image data (p. 695, right column “…the compression performed in the YUV domain…”). It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to apply the teachings and training scheme model of Besenbruch, Yoshimura, Kang, Saripalli, and Dutta to the YUV image data of Mancuso in order to provide a YUV teacher image encoder trained to map YUV images into a compressed domain. Doing so would preserve an image’s luminance quality (see Mancuso p. 697, left column, 10- 12 lines from the bottom, “the sole unquestionable advantage of YUV 422 compression is the luminance preservation”), while decreasing bandwidth and storage requirements. All references aim to improve image processing techniques involving encoding, compressing images, learned feature representations, latent representations, or teacher-student models with a focus on maintaining or improving image quality. Claim(s) 9 and 18 are rejected under 35 U.S.C. 103 as being obvious over Besenbruch et al. (US 20230154055 A1; provided by applicant in IDS) in view of Yoshimura et al. ("DynamicISP: Dynamically Controlled Image Signal Processor for Image Recognition"; copy provided by Examiner), and in further view of Amura et al. (US 20040071210 A1; “Description of the Prior Art” section relied upon by Examiner). Regarding claims 9 and 18, Besenbruch and Yoshimura teach the method of claim 1, and the electronic device of claim 11, including mapping the raw image to a compressed domain. Yoshimura further teaches : processing raw image video sequences and determining image signal processor (ISP) parameters for the first frame and subsequent frames, teaching that ISP parameters may be predicted for a following frame and that parameter initializer determines ISP parameters for the first frame based on another frame (p. 2, left column, lines 7-10 “our method predicts appropriate ISP parameters for the following frame based on intermediate features of the downstream recognition model.”; p. 5, Figure 3 description “The parameter initializer decides the ISP parameters for the first frame based on what kind of parameters were used for the second frames.”; p. 4, right column, lines 2-3 of section 3.2 “The first frame is input to the differentiably implemented ISP…”) Accordingly, Yoshimura teaches a multi-frame stream including a first frame comprising raw image data and subsequent frames that are processed in relation to one another. Besenbruch and Yoshimura fail to explicitly disclose: wherein mapping the raw image to the compressed domain comprises: mapping a multi-frame stream of data including the raw image to the compressed domain based on (i) a first frame within the multi-frame stream of data and (ii) at least one of: spatial redundancy between the first frame and a remainder of frames within the multi-frame stream of data; or the spatial redundancy between the first frame and the remainder of frames and temporal redundancy between consecutive frames within the multi-frame stream of data; and wherein the first frame includes the raw image. In a related art, Amara’s “Description of the Prior Art” section teaches: mapping a multi-frame stream into a compressed representation based on spatial and temporal redundancies among the frames ([0004] “Video data is compressed or coded for transmission by taking advantage of the spatial redundancies within a given frame and the temporal redundancies between successive frames… interframe compression exploits both spatial and temporal redundancies.”). Amara’s cited teachings being found in the “Description of the Prior Art” section is indicative of these techniques being well known in the art prior to the filing date of the claimed invention. It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to apply the known multi-frame compression techniques taught by Amara to the raw image frame sequence of Yoshimura when mapping the raw image information into the compressed domain previously taught by the combination of Besenbruch and Yoshimura (see claim 1 rejection above) because exploiting spatial and temporal redundancies among image frames predictably reduces redundant information, improves compression efficiency, and transmission efficiency (see Amara [0004] “Video data compressions systems play a key role in increasing the efficiency of video data transmission.”). Accordingly, the combination teaches mapping a multi-frame stream of data including a raw image to a compressed domain based on a first frame within the multi-frame stream and at least one of spatial redundancy and/or temporal redundancy among frames of the multi-frame stream, wherein the first frame includes the raw image. Claim(s) 10 and 19 are rejected under 35 U.S.C. 103 as being obvious over Besenbruch et al. (US 20230154055 A1; provided by applicant in IDS) in view of Yoshimura et al. ("DynamicISP: Dynamically Controlled Image Signal Processor for Image Recognition"; copy provided by Examiner), in further view of Zeng et al. (US 20040062448 A1), and in further view of Pope (US 20210224575 A1; provided by applicant in IDS). Regarding claims 10 and 19, Besenbruch and Yoshimura teach the method of claim 1, and the electronic device of claim 11. Besenbruch and Yoshimura fail to explicitly disclose: wherein at least one of: mapping the raw image to the compressed domain comprises ordering data from most significant bits to least significant bits; or performing the one or more image signal processing operations comprises processing only the most significant bits of data in the compressed domain to perform at least one of the one or more image signal processing operations that does not require full-resolution data. In a related art, Zeng teaches: mapping the raw image to the compressed domain comprises ordering data from most significant bits to least significant bits ([0019] “Each sub-band may be considered to be a sequence of binary …known as bitplanes. The first bitplane 106 comprises the array of the most significant bit (MSB) .... The second bitplane 108 comprises the array of the next most significant bit and so forth with the final bitplane 110 comprising the least significant bits (LSB) of the indices. The bit stream is encoded by scanning the values of the bits making up the successive bitplanes.”) It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to map the raw image into Besenbruch and Yoshimuras’ compressed domain using a significance-based ordering from most significant bits to least significant bits as taught by Zeng because Zeng teaches generating compressed representation in which image information is encoded in successive bitplanes from the most significant bits to the least significant bits. Doing so would enable the model to prioritize image information having the greatest impact on image quality while allowing lower significant information to be disregarded when appropriate (see second to last sentence of Zeng [00019]) , thereby increasing compression processing efficiency by reducing redundant processing of less significant bits while maintaining image quality during compression processing. Besenbruch, Yoshimura, and Zeng remain to fail explicitly disclosing: performing the one or more image signal processing operations comprises processing only the most significant bits of data in the compressed domain to perform at least one of the one or more image signal processing operations that does not require full-resolution data. In a related art, Pope teaches: performing image processing using only the most significant portion of image data, teaching that “the 8 most significant bits (MSBs) of input data … may be copied over… as 8-bit output data for further processing” and that the processing circuitry may “receive 16-bit input data and provide 8-bit output data for further processing” (Pope [0072]). Accordingly, Pope teaches performing an image processing operation using only the most significant bits of image data without requiring the full-resolution data. It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to perform the image signal processing (ISP) taught by Besenbruch and Yoshimura (in claim 1 rejection) on only the most significant portions of Besenbruch and Yoshimura’s compressed domains because Pope teaching that image processing operations may be performed using only the most significant bits of image data rather than the full-resolution data. Doing so would reduce the amount of data processed while retaining information for image processing, thereby increasing the model’s efficiency while maintaining image quality of the most significant quality portions of the image data. All four references deal teach techniques for reducing the amount of image data that must be processed while retaining image data sufficient for compression, reconstruction, and image processing. Claim 20 is rejected under 35 U.S.C. 103 as being obvious over Dutta (US 20210265018 A1) in view of Besenbruch et al. (US 20230154055 A1; provided by applicant in IDS), and in further view of Yoshimura et al. ("DynamicISP: Dynamically Controlled Image Signal Processor for Image Recognition"; copy provided by Examiner). Regarding claim 20, Dutta teaches: A non-transitory machine readable medium containing instructions that when executed cause at least one processor of an electronic device to ([0221] “a non-transitory computer readable storage medium storing instructions executable by a processor to perform any of the methods described…”; [0178]) : The remaining limitations of claim 20 equally mirror the scope of claims 1 and 11 taught by Besenbruch and Yoshimura, found above. For sake of brevity, refer back to the teachings of Besenbruch and Yoshimura found in the 103 rejections for claims 1 and 11. Besenbruch and Yoshimura failed to teach a non-transitory machine readable medium containing instructions that when executed cause at least one processor of an electronic device. It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to store the instructions for performing the compression model for image signal processing taught by Besenbruch and Yoshimura (seen in 103 rejections for claims 1 and 11) on the non-transitory computer machine readable medium taught by Dutta because using a non-transitory computer machine readable medium to store instructions implementing a known computer implemented method is a predictable variation that yields no unexpected result and merely places a known method into known software-storage format. All three inventions use neural networks to increase the speed of processing images with the goal of increasing and/or maintaining image quality, while Dutta and Besenbruch focus on compression techniques for imaging. Conclusion Any inquiry concerning this communication or earlier communications from the examiner should be directed to SAMUEL DAVID BAYNES whose telephone number is (571)272-0607. The examiner can normally be reached Monday - Friday 8:00 am - 5:00 pm. 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. /SDB/ Samuel Baynes Examiner, Art Unit 2665 /Stephen R Koziol/Supervisory Patent Examiner, Art Unit 2665 Application/Control Number: 18/816,846 Page 2 Art Unit: 2665 Application/Control Number: 18/816,846 Page 3 Art Unit: 2665 Application/Control Number: 18/816,846 Page 4 Art Unit: 2665 Application/Control Number: 18/816,846 Page 5 Art Unit: 2665 Application/Control Number: 18/816,846 Page 6 Art Unit: 2665 Application/Control Number: 18/816,846 Page 7 Art Unit: 2665 Application/Control Number: 18/816,846 Page 8 Art Unit: 2665 Application/Control Number: 18/816,846 Page 9 Art Unit: 2665 Application/Control Number: 18/816,846 Page 10 Art Unit: 2665 Application/Control Number: 18/816,846 Page 11 Art Unit: 2665 Application/Control Number: 18/816,846 Page 12 Art Unit: 2665 Application/Control Number: 18/816,846 Page 13 Art Unit: 2665 Application/Control Number: 18/816,846 Page 14 Art Unit: 2665 Application/Control Number: 18/816,846 Page 15 Art Unit: 2665 Application/Control Number: 18/816,846 Page 16 Art Unit: 2665 Application/Control Number: 18/816,846 Page 17 Art Unit: 2665 Application/Control Number: 18/816,846 Page 18 Art Unit: 2665 Application/Control Number: 18/816,846 Page 19 Art Unit: 2665 Application/Control Number: 18/816,846 Page 20 Art Unit: 2665 Application/Control Number: 18/816,846 Page 21 Art Unit: 2665 Application/Control Number: 18/816,846 Page 22 Art Unit: 2665 Application/Control Number: 18/816,846 Page 23 Art Unit: 2665
Prosecution Insights