Abstract: An example device for decoding video data includes a memory configured to store video data and one or more processors implemented in circuitry and configured to determine a maximum possible value for a secondary transform syntax element for a block of video data, entropy decode a value for the secondary transform syntax element of the block to form a binarized value representative of the secondary transform for the block, reverse binarize the value for the secondary transform syntax element using a common binarization scheme regardless of the maximum possible value to determine the secondary transform for the block, and inverse-transform transform coefficients of the block using the determined secondary transform.

Abstract: A video encoder may determine a set of quantization offset parameters for a group of scaled transform coefficients for a block of video data based on side information associated with the block of video data. The video encoder may further quantize the group of scaled transform coefficients for the block of video data to generate quantized transform coefficients for the block of video data based at least in part on the set of quantization offset parameters. The video encoder may further generate an encoded video bitstream based at least in part on the quantized transform coefficients for the block of video data.

Abstract: An example method of decoding video data includes obtaining, from a coded video bitstream and for a current block of the video data, an indication of an intra-prediction mode that identifies an initial predictive block; filtering, in parallel, samples in a current line of a plurality of lines of the initial predictive block based on filtered values of samples in a preceding line of the plurality of lines and unfiltered values of samples in the current line to generate filtered values for samples for the current line; and reconstructing, using intra prediction, values of samples of the current block based on the filtered values of the samples of the current initial predictive block and residual data for the current block that represents a difference between the filtered values of the samples of the current initial predictive block and the values of samples of the current block.

Abstract: Systems and techniques are described for encoding and/or decoding data based on motion estimation that applies variable-scale warping. An encoding device can receive an input frame and a reference frame that depict a scene at different times. The encoding device can generate an optical flow identifying movements in the scene between the two frames. The encoding device can generate a weight map identifying how finely or coarsely the reference frame can be warped for input frame prediction. The encoding device can generate encoded video data based on the optical flow and the weight map. A decoding device can generate a reconstructed optical flow and a reconstructed weight map from the encoded data. A decoding device can generate a prediction frame by warping the reference frame based on the reconstructed optical flow and the reconstructed weight map. The decoding device can generate a reconstructed input frame based on the prediction frame.

Abstract: A method of image compression includes receiving an image. Multiple quantized latent representations are generated to represent features of the image. Each of the quantized latent representations has a different resolution and is generated at staggered timings. Each of the later generated quantized latent representations is conditioned on each of the prior generated quantized latent representations. The multiple quantized latent representations are decoded to reconstruct the image.

Abstract: A video decoder can be configured to receive, in a syntax structure that applies to a current block, graph-related information; determine a transform matrix based on the received graph-related syntax information; perform an inverse transform based on the determined transform matrix of one or more coefficient values to generate a residual block; and reconstruct the current block of the video data based on the residual block.

Abstract: Techniques are described of hybrid coders that are configured to selectively use adaptive or non-adaptive coding techniques. A video coder (e.g., video encoder or video decoder) may code (e.g., encode or decode) first video data (e.g., a syntax element or value), for coding a first block, based on an adaptive context model (e.g., one or more adaptive context models) and code second video data, for coding a second block, based on a non-adaptive context model (e.g., one or more non-adaptive context models).

Abstract: Certain aspects of the present disclosure provide techniques for compressing content using a neural network. An example method generally includes receiving content for compression. The content is encoded into a first latent code space through an encoder implemented by an artificial neural network trained to generate a latent space representation of the content. A first compressed version of the encoded content is generated using a first quantization bin size of a series of quantization bin sizes. A refined compressed version of the encoded content is generated by scaling the first compressed version of the encoded content into one or more second quantization bin sizes smaller than the first quantization bin size, conditioned at least on a value of the first compressed version of the encoded content. The refined compressed version of the encoded content is output for transmission.

Abstract: Techniques are described for learned bidirectional predicted frame (B-frame) coding. An example method can include receiving a residual associated with a frame of a current time step; determining first motion information for a first reference frame associated with a first time step and second motion information for a second reference frame associated with a second time step, wherein the current time step is after the first time step and before the second time step; determining third motion information for the frame based on the first motion information and second motion information; generating a predicted frame based on the third motion information, first reference frame and second reference frame; and generating, using the predicted frame and residual, a reconstructed B-frame for the current time step, the reconstructed B-frame representing the frame.

Abstract: A computer-implemented method for operating an artificial neural network (ANN) includes receiving an input by the ANN. The ANN generates a latent representation of the input. The latent representation is communicated according to a bit rate based on a learned latent scaling parameter. The latent scaling parameter is learned based on a channel index and a tradeoff parameter value that corresponds to a value that balances the bit rate and a distortion.

Abstract: Techniques are described herein for video coding, including compression of bitstream indexes for neural network based video coding and/or parallel entropy coding. One example includes obtaining a sequence of video data, identifying positions in the sequence of video data associated with entry points for individually entropy codable parcels of a parallel entropy codable sequence of video data, and generating the parallel entropy codable sequence of video data. An index is then generated for the parallel entropy codable sequence of video data, the index identifying the individually entropy codable parcels within the parallel entropy codable sequence of video data.

Abstract: A video decoder may receive, in a bitstream that comprises an encoded representation of video data, information indicating whether a residual block is partitioned and information indicating a partition tree type for the residual block based on the residual block being partitioned, wherein the residual block is indicative of a difference between a current block and a prediction block. The video decoder may determine, based on the received information that the residual block is partitioned and the partition tree type for the residual block, a plurality of residual sub-blocks into which the residual block is partitioned according to the partition tree type. The video decoder may produce the residual data for the current block based at least in part on the residual block being partitioned according to the partition tree type into the plurality of residual sub-blocks and may decode the current block using the residual data.

Abstract: Techniques are described herein for processing video data. For instance, a process can include obtaining encoded video data. The process can include determining an intersection of values between values for a first termination byte of a first parcel of the encoded video data and values of a second termination byte of a second parcel of the encoded video data. The process can further include determining a joint termination byte for the first termination byte of the first parcel and the second termination byte of the second parcel. Values for the joint termination byte are based on the intersection of values. The process can include generating entropy coded data including the joint termination byte for the first parcel and the second parcel. The entropy coded data can be generated using arithmetic coding or binary coding.

Abstract: This disclosure describes examples of extending the number of available discrete cosine transform (DCT) and discrete sine transform (DST) for encoding and decoding. A video coder may determine one or more transforms or inverse transforms to apply from a set of transforms or inverse transforms that includes DCT-2 or inverse DCT-2, DST-7 or inverse DST-7, DST-8 or inverse DST-8, DCT-3 or inverse DCT-3, DST-2 or inverse DST-2, DST-3 or inverse DST-3, DCT-4 or inverse DCT-4, DST-4 or inverse DST-4, DST-5 or inverse DST-5, DST-6 or inverse DST-6, and identity transform an inverse identity transform (IDT).

Abstract: An example device for coding video data includes a memory configured to store video data; and one or more processors implemented in circuitry and configured to: code a first codeword representing a selected transform scheme of a set of transform candidates of a multiple transform selection (MTS) scheme for a current block of video data, the selected transform scheme being a secondary transform of a set of available secondary transforms to be applied in addition to a primary transform; code a second codeword representing the secondary transform from the set of available secondary transforms; and apply the primary transform and the secondary transform during coding of residual data for the current block. The second codeword may be a value for a low-frequency non-separable transform (LFNST) syntax element.

Abstract: An example device for coding video data includes a memory configured to store video data; and one or more processors implemented in circuitry and configured to: code a first codeword representing a selected transform scheme of a set of transform candidates of a multiple transform selection (MTS) scheme for a current block of video data, the selected transform scheme being a secondary transform of a set of available secondary transforms to be applied in addition to a primary transform; code a second codeword representing the secondary transform from the set of available secondary transforms; and apply the primary transform and the secondary transform during coding of residual data for the current block. The second codeword may be a value for a low-frequency non-separable transform (LFNST) syntax element.

Abstract: Techniques are described for improving transform coding. For example, an encoded block of video data can be obtained, and a width and/or a height of the block can be determined. The width can be compared to a first threshold and/or the height can be compared to a second threshold. A horizontal transform and a vertical transform can be determined for the block based on comparing the width of the block to the first threshold and/or the height of the block to the second threshold. The horizontal transform and the vertical transform are determined without decoding a syntax element that indicates the horizontal transform and the vertical transform (e.g., the syntax element is not in an encoded video bitstream processed by a decoding device). In some cases, residual data is determined using the horizontal and vertical transforms, and a video block is determined using the residual data and a predictive block.

Abstract: An example device for decoding video data includes a memory configured to store the video data and one or more processors coupled to the memory. The one or more processors are configured to reorganize 2-D dequantized coefficients according to a first ordering. The one or more processors are configured to apply an inverse low-frequency non-separable transform (LFNST) to the reorganized 2-D dequantized coefficients to create inverse transformed coefficients. The one or more processors are configured to reorganize the inverse transformed coefficients according to a second ordering, the second ordering being based on an array including values, wherein each value in the array corresponds to a position in a 2-D block and the values in the array denote indices of the 2-D block in a defined order. The one or more processors are configured to decode the video data based on the second ordered inverse transformed coefficients.

Abstract: An example method for entropy decoding of video data includes retrieving a pre-defined initialization value for a context of a plurality of contexts used in a context-adaptive entropy coding process to entropy code a value for a syntax element for an independently codable unit of video data; determining, based on the pre-defined initialization value and in a linear domain, an initial probability state of the context; and entropy decoding, from a bitstream and based on the initial probability state of the context, a bin of the value for the syntax element.

Abstract: A video encoder determines scaled transform coefficients, wherein determining the scaled transform coefficients comprises scaling transform coefficients of a block of the video data according to a given quantization step. The video encoder determines scalar quantized coefficients, wherein determining the scalar quantized coefficients comprises applying scalar quantization to the scaled transform coefficients of the block. Additionally, the video encoder applies a neural network that determines a respective set of probabilities for each respective transform coefficient of the block. The respective set of probabilities for the respective transform coefficient includes a respective probability value for each possible adjustment value in a plurality of possible adjustment values. Inputs to the neural network include the scaled transform coefficients and the scalar quantized coefficients.