Abstract: A method for decoding point cloud data comprises, based on a comparison of a maximum difference value and a threshold, applying an inverse function to a set of one or more jointly coded values to recover (i) residual values for attribute values of a current point of point cloud data and (ii) a predictor index that indicates a predictor in a predictor list, wherein predictors in the predictor list are based on attribute values of one or more neighbor points; determining predicted attribute values based on the predictor index; and reconstructing the attribute values of the current point based on the residual values and the predicted attribute values.

Abstract: A device comprises one or more processors configured to: obtain a value for a first laser, the value for the first laser indicating a number of probes in an azimuth direction of the first laser; decode a syntax element for a second laser, wherein the syntax element for the second laser indicates a difference between the value for the first laser and a value for the second laser, the value for the second laser indicating a number of probes in the azimuth direction of the second laser; determine the value for the second laser indicating the number of probes in the azimuth direction of the second laser based on the first value and the indication of the difference between the value for the first laser and the value for the second laser; and decode a point based on the number of probes in the azimuth direction of the second laser.

Abstract: Certain aspects of the present disclosure provide techniques for global motion modeling. Embodiments include receiving a first image and a second image from a camera attached to a moving object. Embodiments include identifying a pixel in the first image. Embodiments include determining, based on one or more parameters associated with the camera, a vector representing a range of locations in which a real-world point corresponding to the pixel is likely to be found in the second image, wherein the parameters associated with the camera comprise: a first parameter related to a location of the camera relative to a ground surface; a second parameter related to motion of the moving object; and a third parameter related to an orientation of the camera relative to the ground surface. Embodiments include determining, using the vector, a location of the real-world point in the second image.

Abstract: A G-PCC encoder and G-PCC decoder may quantize and scale, respectively, a position of a child node. The G-PCC encoder may control the precision of the quantization and scaling using a quantization parameter (QP) value and a parameter value k, wherein the parameter value k specifies a number of QP points per doubling of a scaling step size to be used at the G-PCC decoder.

Abstract: Disclosed are systems, apparatuses, processes, and computer-readable media to detect objects in dynamic lighting conditions. A method of processing image data includes obtaining, at a computing device, a first image of an object at a first position in an environment, obtaining, at the computing device, a second image of the object at a second position in the environment, and determining, at the computing device, movement of the object in the first image and the second image at least in part using an optical flow engine, wherein the optical flow engine is trained based on augmented training data generated using at least one of noise associated with low ambient lighting conditions, noise associated with motion blur due to exposure of an image sensor in low ambient lighting conditions, or brightness variations. The object and/or the computing device may move between the first image and the second image.

Abstract: A method of decoding video data includes determining an intra prediction mode from a plurality of intra prediction modes for a current block of the video data, determining a low frequency non-separable transform (LFNST) kernel from a plurality of LFNST kernels for the current block based on the determined intra prediction mode, wherein at least one LFNST kernel of the plurality of LFNST kernels is associated with at least two different intra prediction modes of the plurality of intra prediction modes, applying an inverse of the determined LFNST kernel to coefficient values generated from a transform unit (TU) of the current block to generate intermediate values, applying an inverse primary transform on the intermediate values to generate residual data, and reconstructing the current block based on the residual data.

Abstract: A G-PCC coder is configured to receive the point cloud data, determine a final quantization parameter (QP) value for the point cloud data as a function of a node QP offset multiplied by a geometry QP multiplier, and code the point cloud data using the final QP value to create an coded point cloud.

Abstract: In some examples, a method of decoding a point cloud includes decoding an initial QP value from an attribute parameter set. The method also includes determining a first QP value for a first component of an attribute of point cloud data from the initial QP value. The method further includes determining a QP offset value for a second component of the attribute of the point cloud data and determining a second QP value for the second component of the attribute from the first QP value and from the QP offset value. The method includes decoding the point cloud data based on the first QP value and further based on the second QP value.

Abstract: A method of decoding point cloud data comprises obtaining a bitstream that includes an arithmetically encoded syntax element indicating a vertical plane position of a planar mode of a node; and decoding the vertical plane position of the planar mode in the node, wherein decoding the vertical plane position of the planar mode comprises: determining a laser index of a laser candidate in a set of laser candidates, wherein the determined laser index indicates a laser beam that intersects the node; determining a context index based on whether the laser beam is above a first distance threshold, between the first distance threshold and a second distance threshold, between the second distance threshold and a third distance threshold, or below the third distance threshold; and arithmetically decoding the vertical plane position of the planar mode using a context indicated by the determined context index.

Abstract: An example device for decoding point cloud data includes memory configured to store the point cloud data and one or more processors implemented in circuitry and coupled to the memory. The one or more processors are configured to determine dimensions of a region box and determine dimensions of a slice bounding box. The one or more processors are also configured to decode a slice of the point cloud data associated with the slice bounding box. The dimensions of the region box are constrained to not exceed the dimensions of the slice bounding box.

Abstract: Techniques for performing inverse transform operations on high bit-depth video data are described. A video decoder may receive encoded video data encoded at a first bit-depth in an encoded video bitstream. The video decoder may determine one or more of a dequantization shift or a mid-transform shift based on information in the encoded video bitstream, and perform an inverse transform on the encoded video data at a second bit-depth using the dequantization shift and the mid-transform shift, wherein the second bit-depth is lower than the first bit-depth.

Abstract: An example device for decoding point cloud data includes memory configured to store the point cloud data and one or more processors implemented in circuitry and coupled to the memory. The one or more processors are configured to determine dimensions of a region box and determine dimensions of a slice bounding box. The one or more processors are also configured to decode a slice of the point cloud data associated with the slice bounding box. The dimensions of the region box are constrained to not exceed the dimensions of the slice bounding box.

Abstract: An example device for processing point cloud data includes a memory configured to store the point cloud data and one or more processors implemented in circuitry and coupled to the memory. The one or more processors are configured to count a number of edges of a cube of point cloud data comprising a vertex. The one or more processors are configured to set a variable based on a total of the counting. The one or more processors are also configured to process the point cloud data based on the variable.

Abstract: A method of decoding point cloud data comprises obtaining a bitstream that includes an arithmetically encoded syntax element indicating a vertical point position offset within a node of a tree that represents 3-dimensional positions of points in a point cloud represented by the point cloud data; and decoding the vertical point position offset, wherein decoding the vertical point position offset comprises: determining a laser index of a laser candidate in a set of laser candidates, wherein the determined laser index indicates a laser beam that intersects the node; determining a context index based on whether the laser beam is above a first distance threshold, between the first distance threshold and a second distance threshold, between the second distance threshold and a third distance threshold, or below the third distance threshold; and arithmetically decoding a bin of the vertical point position offset using a context indicated by the determined context index.

Abstract: In some examples, a method of decoding a point cloud includes decoding an initial QP value from an attribute parameter set. The method also includes determining a first QP value for a first component of an attribute of point cloud data from the initial QP value. The method further includes determining a QP offset value for a second component of the attribute of the point cloud data and determining a second QP value for the second component of the attribute from the first QP value and from the QP offset value. The method includes decoding the point cloud data based on the first QP value and further based on the second QP value.

Abstract: In some examples, a method of decoding a point cloud includes determining a first slice QP value for a first component of an attribute in a slice of point cloud data. The method also includes decoding a first delta QP value for the first component of the attribute for a region in the slice and determining a first region QP value for the first component of the attribute in the region from the first slice QP value and from the first delta QP value. The method further includes decoding a second delta QP value for the second component of the attribute for the region and determining a second region QP value for the second component of the attribute in the region from the second delta QP value. The method includes decoding the point cloud data based on the first and second region QP values.

Abstract: A device for decoding video data receives encoded video data encoded at a first bit-depth; receives in an encoded video bitstream for the video data one or more syntax elements; determines a dequantization shift and a mid-transform shift based on the one or more syntax elements; and performs an inverse transform of the encoded video data at a second bit-depth using the dequantization shift and the mid-transform shift to determine residual values for a block of video data, wherein the second bit-depth is different than the first bit-depth.

Abstract: Techniques for performing inverse transform operations on high bit-depth video data are described. A video decoder may receive encoded video data encoded at a first bit-depth in an encoded video bitstream. The video decoder may determine one or more of a dequantization shift or a mid-transform shift based on information in the encoded video bitstream, and perform an inverse transform on the encoded video data at a second bit-depth using the dequantization shift and the mid-transform shift, wherein the second bit-depth is lower than the first bit-depth.

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 size of a current block of video data; determine an intra-prediction mode for the current block of video data; determine a mode group including the determined intra-prediction mode, the mode group being one of a plurality of mode groups, each including respective sets of intra-prediction modes; determine a set of available multiple transform selection (MTS) schemes for the current block according to the size and the intra-prediction mode for the current block; determine an MTS scheme from the set of available MTS schemes according to the determined mode group; apply transforms of the MTS scheme to a transform block of the current block to produce a residual block for the current block; and decode the current block using the residual block.

Abstract: A method of decoding video data includes determining an intra prediction mode from a plurality of intra prediction modes for a current block of the video data, determining a low frequency non-separable transform (LFNST) kernel from a plurality of LFNST kernels for the current block based on the determined intra prediction mode, wherein at least one LFNST kernel of the plurality of LFNST kernels is associated with at least two different intra prediction modes of the plurality of intra prediction modes, applying an inverse of the determined LFNST kernel to coefficient values generated from a transform unit (TU) of the current block to generate intermediate values, applying an inverse primary transform on the intermediate values to generate residual data, and reconstructing the current block based on the residual data.