Abstract: Artificial image generation may include obtaining a source image, identifying quantization information from the source image, wherein identifying the quantization information includes identifying multiresolution quantization interval information from the source image, generating a restoration filtered image by restoration filtering the source image, generating a constrained restoration filtered image by constraining the restoration filtered image based on the quantization information, obtaining an unconstrained artificial image based on the constrained restoration filtered image and a generative artificial neural network obtained using a generative adversarial network, obtaining the artificial image by constraining the unconstrained artificial image based on the quantization information, and outputting the artificial image.

Abstract: Spiking neural networks that perform temporal encoding for phase-coherent neural computing are provided. In particular, according to an aspect of the present disclosure, a spiking neural network can include one or more spiking neurons that have an activation layer that uses a double exponential function to model a leaky input that an incoming neuron spike provides to a membrane potential of the spiking neuron. The use of the double exponential function in the neuron's temporal transfer function creates a better defined maximum in time. This allows very clearly defined state transitions between “now” and the “future step” to happen without loss of phase coherence.

Abstract: A spatially adaptive quantization-aware deblocking filter is used for encoding or decoding video or image frames. The deblocking filter receives a reconstructed frame produced based on dequantized and inverse transformed coefficients of a video frame or an image frame. The reconstructed frame is filtered according to adaptive quantization field data for the video or image frame. The adaptive quantization field data represents weights applied to quantization values used at different areas of the video or image frame. A number of blocking artifacts remaining within the resulting filtered frame is determined. The adaptive quantization field data is then adjusted based on that number of blocking artifacts. The filtered frame is then filtered according to the adjusted adaptive quantization field data. The resulting re-filtered frame is then output to an output source, such as for transmission, display, storage, or further processing.

Abstract: An image encoder includes a processor and a memory. The memory includes instructions configured to cause the processor to perform operations. In one example implementation, the operations may include determining whether a dictionary item is available for replacing a block of an image being encoded, the determining based on a hierarchical lookup mechanism, and encoding the image along with reference information of the dictionary item in response to determining that the dictionary item is available. In one more example implementation, the operations may include performing principal component analysis (PCA) on a block to generate a corresponding projected block, the block being associated with a group of images, comparing the projected block with a corresponding threshold, descending the block recursively based on the threshold until a condition is satisfied, and identifying a left over block as a cluster upon satisfying of the condition.

Abstract: The loss of image quality during compression is controlled using a sequence of quality control metrics. The sequence of quality control metrics is selected for quantizing transform coefficients within an area of the image based on an error level definition. Candidate bit costs are then determined by quantizing the transform coefficients according to the error level definition or a modified error level and the sequence of quality control metrics. Where the candidate bit cost resulting from using the modified error level is lower than the candidate bit cost resulting from using the error level definition, the transform coefficients are quantized according to the modified error level and the sequence of quality control metrics. Otherwise, the transform coefficients are quantized based on the error level definition and according to the sequence of quality control metrics.

Abstract: An apparatus includes a processor that is configured to select a palette entry in the palette for coding a value of a pixel of the image block; obtain respective palette indexes of neighboring pixels of the pixel; select, using the respective palette indexes, an entropy code for coding an index of the palette entry; and code the palette entry using the entropy code. A method includes obtaining respective palette indexes of neighboring pixels of a pixel of the image block; selecting an entropy code using the respective palette indexes; decoding, from a encoded bitstream, an index of a palette entry; and setting a value of the pixel using the palette entry.

Abstract: An image encoder includes a processor and a memory. The memory includes instructions configured to cause the processor to perform operations. In one example implementation, the operations may include determining whether a dictionary item is available for replacing a block of an image being encoded, the determining based on a hierarchical lookup mechanism, and encoding the image along with reference information of the dictionary item in response to determining that the dictionary item is available. In one more example implementation, the operations may include performing principal component analysis (PCA) on a block to generate a corresponding projected block, the block being associated with a group of images, comparing the projected block with a corresponding threshold, descending the block recursively based on the threshold until a condition is satisfied, and identifying a left over block as a cluster upon satisfying of the condition.

Abstract: The present disclosure provides systems and methods for compressing and/or distributing machine learning models. In one example, a computer-implemented method is provided to compress machine-learned models, which includes obtaining, by one or more computing devices, a machine-learned model. The method includes selecting, by the one or more computing devices, a weight to be quantized and quantizing, by the one or more computing devices, the weight. The method includes propagating, by the one or more computing devices, at least a part of a quantization error to one or more non-quantized weights and quantizing, by the one or more computing devices, one or more of the non-quantized weights. The method includes providing, by the one or more computing devices, a quantized machine-learned model.

Abstract: A method for generating random numbers includes initializing a pseudo-random number generator (PRNG) having a state of 2048 bits comprising inner bits and outer bits, the inner bits comprising the first 128 bits of the 2048 bits and the outer bits comprising the remaining bits of the 2048 bits. The method also includes retrieving AES round keys from a key source, and for a threshold number of times, executing a round function using the AES round keys by XOR'ing odd-numbered branches of a Feistel network having 16 branches of 128 bits with a function of corresponding even-numbered neighbor branches of the Feistel network, and shuffling each branch of 128 bits into a prescribed order. The method also includes executing an XOR of the inner bits of the permuted state with the inner bits of a previous state.

Abstract: Artificial image generation may include obtaining a source image, identifying quantization information from the source image, wherein identifying the quantization information includes identifying multiresolution quantization interval information from the source image, generating a restoration filtered image by restoration filtering the source image, generating a constrained restoration filtered image by constraining the restoration filtered image based on the quantization information, obtaining an unconstrained artificial image based on the constrained restoration filtered image and a generative artificial neural network obtained using a generative adversarial network, obtaining the artificial image by constraining the unconstrained artificial image based on the quantization information, and outputting the artificial image.

Abstract: Encoding using locally mixed colors is disclosed. A method for encoding an image block using palletization includes selecting a fixed palette for the image block, the fixed palette including fixed palette entries; selecting a mixed palette for the image block, the mixed palette including mixed palette entries, each mixed palette entry corresponding, respectively, to a pixel neighborhood, a mixing of the pixel neighborhood, and a manipulation of the mixing of the pixel neighborhood; determining a pixel map, the pixel map comprising, for a pixel of at least some pixels of the image block, a respective mapping to one of a fixed palette entry or a mixed palette entry; and encoding, in an encoded bitstream, the pixel map.

Abstract: An image block is coded using entropy-inspired directional filtering. During encoding, intensity differences are determined for at least some pixels of an image block based on neighboring pixels of respective ones of the at least some pixels. Angles are estimated for each of those pixels based on the intensity differences. A main filtering direction of the image block is then determined based on the estimated angles. The image block is filtered according to the main filtering direction to remove artifacts along oblique edges associated with the image block. The filtered image block is then encoded to an encoded image. During decoding, an angular map indicating angles estimated for pixels of an encoded image block is received and used to determine the main filtering direction of the image block. The image block can then be filtered according to the main filtering direction and then output for display or storage.

Abstract: A method includes obtaining respective filtered pixels for pixels of a reconstructed image; and obtaining an edge-preserved image using the respective filtered pixels. Obtaining the respective filtered pixels includes, for each pixel of the reconstructed image, obtaining a respective filtered pixel by selecting a pixel patch including the pixel and first neighboring pixels of the pixel that are at relative neighboring locations with respect to the pixel; calculating respective weights for the first neighboring pixels; and filtering the pixel using the respective weights of the first neighboring pixels and the neighboring pixels to obtain the respective filtered pixel.

Abstract: Techniques of compressing color images in the presence of chromatic aberrations involve performing, prior to compression, a chromatic aberration correction operation on a color image. Along these lines, the chromatic aberration of an imaging system may be represented as a vector displacement map between a red channel and a green channel of a color image, a blue channel and a green channel of the color image, or both. In some implementations, prior to adding the vector displacements to each of the images of the red channel and the blue channel, these displacements are weighted according to proximity from an edge of each of the respective red and blue images. In some further implementations, the vector displacement maps as well as the weights are blurred with a blurring kernel such as a gaussian. Once these vector displacements are added to each of the red and blue images, the resulting color images are linearly combined to produce a new brightness channel Y.

Abstract: A method for decoding an image block includes receiving a quantized transform block; generating a decoded block from the quantized transform block; applying an edge-preserving filter to the decoded block, to obtain an edge-filtered decoded block; obtaining a transformed edge-preserved block using a transform type and quantization data to the edge-filtered decoded block; clamping a value of the transformed edge-preserved block to a corresponding value of the quantized transform block to obtain a smoothed transform block; and inverse transforming the smoothed transform block to obtain an edge-preserved smoothed block. The applying an edge-preserving filter to the decoded block includes determining respective patch-based distances between a pixel of the decoded block and neighboring pixels; determining respective weights corresponding to the respective patch-based distances; and filtering the pixel using the respective weights and the neighboring pixels.

Abstract: A method for decoding an image block includes receiving a quantized transform block; generating a decoded block from the quantized transform block; applying an edge-preserving filter to the decoded block, to obtain an edge-filtered decoded block; obtaining a transformed edge-preserved block using a transform type and quantization data to the edge-filtered decoded block; clamping a value of the transformed edge-preserved block to a corresponding value of the quantized transform block to obtain a smoothed transform block; and inverse transforming the smoothed transform block to obtain an edge-preserved smoothed block. The applying an edge-preserving filter to the decoded block includes determining respective patch-based distances between a pixel of the decoded block and neighboring pixels; determining respective weights corresponding to the respective patch-based distances; and filtering the pixel using the respective weights and the neighboring pixels.

Abstract: Systems, methods, and computer readable media related to generating query variants for a submitted query. In many implementations, the query variants are generated utilizing a generative model. A generative model is productive, in that it can be utilized to actively generate a variant of a query based on application of tokens of the query to the generative model, and optionally based on application of additional input features to the generative model.

Abstract: Morphological anti-ringing is disclosed. A method for encoding a block of an image includes generating a transform block for the block of first values, generating quantized transform coefficients for the transform block, determining a clamping value for the block, encoding the quantized transform coefficients in a compressed bitstream, and encoding the clamping value in the compressed bitstream. The clamping value is used to clamp second values. The second values correspond to the first values and result from inverse transforming the transform block. An apparatus for decoding a block of an image includes a memory and a processor configured to execute instructions stored in the memory to decode, from a compressed bitstream, a quantized transform block; obtain, using the quantized transform block, a decoded block of pixel values; decode, from a compressed bitstream, a clamping value for the block; and clamp the pixel values using the clamping value.

Abstract: An image block is coded using entropy-inspired directional filtering. During encoding, intensity differences are determined for at least some pixels of an image block based on neighboring pixels of respective ones of the at least some pixels. Angles are estimated for each of those pixels based on the intensity differences. A main filtering direction of the image block is then determined based on the estimated angles. The image block is filtered according to the main filtering direction to remove artifacts along oblique edges associated with the image block. The filtered image block is then encoded to an encoded image. During decoding, an angular map indicating angles estimated for pixels of an encoded image block is received and used to determine the main filtering direction of the image block. The image block can then be filtered according to the main filtering direction and then output for display or storage.

Abstract: A spatially adaptive quantization-aware deblocking filter is used for encoding or decoding video or image frames. The deblocking filter receives a reconstructed frame produced based on dequantized and inverse transformed coefficients of a video frame or an image frame. The reconstructed frame is filtered according to adaptive quantization field data for the video or image frame. The adaptive quantization field data represents weights applied to quantization values used at different areas of the video or image frame. A number of blocking artifacts remaining within the resulting filtered frame is determined. The adaptive quantization field data is then adjusted based on that number of blocking artifacts. The filtered frame is then filtered according to the adjusted adaptive quantization field data. The resulting re-filtered frame is then output to an output source, such as for transmission, display, storage, or further processing.