摘要:
Multi-threaded implementations of deblock filtering improve encoding and/or decoding efficiency. For example, a video encoder or decoder partitions a video picture into multiple segments. The encoder/decoder selects between multiple different patterns for splitting operations of deblock filtering into multiple passes. The encoder/decoder organizes the deblock filtering as multiple tasks, where a given task includes the operations of one of the passes for one of the segments. The encoder/decoder then performs the tasks with multiple threads. The performance of the tasks is constrained by task dependencies which, in general, are based at least in part on which lines of the picture are in the respective segments and which deblock filtering operations are in the respective passes. The task dependencies can include a cross-pass, cross-segment dependency between a given pass of a given segment and an adjacent pass of an adjacent segment.
摘要:
Techniques and tools for interpolation of image/video content are described. For example, a tool such as a display processing module in a computing device receives pixel values of a low-resolution picture and determines an interpolated pixel value between a set of the pixel values from the low-resolution picture. The tool uses auto-regressive edge-directed interpolation that incorporates a backward projection constraint (AR-EDIBC). As part of the AR-EDIBC, the tool can compute auto-regressive (AR) coefficients then apply the AR coefficients to the set of pixel values to determine the interpolated pixel value. For the backward projection constraint, the tool accounts for effects of projecting interpolated pixel values back to the pixel values of the low-resolution picture. The tool stores the interpolated pixel values and pixel values from the low-resolution picture as part of a high-resolution picture. The tool can adaptively use AR-EDIBC depending on content and other factors.
摘要:
Multi-threaded implementations of deblock filtering improve encoding and/or decoding efficiency. For example, a video encoder or decoder partitions a video picture into multiple segments. The encoder/decoder selects between multiple different patterns for splitting operations of deblock filtering into multiple passes. The encoder/decoder organizes the deblock filtering as multiple tasks, where a given task includes the operations of one of the passes for one of the segments. The encoder/decoder then performs the tasks with multiple threads. The performance of the tasks is constrained by task dependencies which, in general, are based at least in part on which lines of the picture are in the respective segments and which deblock filtering operations are in the respective passes. The task dependencies can include a cross-pass, cross-segment dependency between a given pass of a given segment and an adjacent pass of an adjacent segment.
摘要:
Techniques and tools for interpolation of image/video content are described. For example, a tool such as a display processing module in a computing device receives pixel values of a low-resolution picture and determines an interpolated pixel value between a set of the pixel values from the low-resolution picture. The tool uses auto-regressive edge-directed interpolation that incorporates a backward projection constraint (AR-EDIBC). As part of the AR-EDIBC, the tool can compute auto-regressive (AR) coefficients then apply the AR coefficients to the set of pixel values to determine the interpolated pixel value. For the backward projection constraint, the tool accounts for effects of projecting interpolated pixel values back to the pixel values of the low-resolution picture. The tool stores the interpolated pixel values and pixel values from the low-resolution picture as part of a high-resolution picture. The tool can adaptively use AR-EDIBC depending on content and other factors.
摘要:
Disclosed herein are representative embodiments of methods, apparatus, and systems for manipulating bitstreams of digital media data compressed according to a compression standard. Also disclosed are representative embodiments of methods, apparatus, and systems for evaluating compliance of an encoded bitstream of digital media data with a compression standard. In one exemplary embodiment, a conforming bitstream of compressed digital media data is input. One or more of the parameters in the bitstream are selectively altered into parameters that do not conform to the video compression standard. The selective alteration can be performed such that parameters that would make the bitstream non-decodable if altered are bypassed and left unaltered. A non-conforming bitstream that includes the one or more selectively altered parameters is output.
摘要:
Techniques and tools for reducing latency in video decoding for real-time communication applications that emphasize low delay. For example, a tool such as a video decoder selects a low-latency decoding mode. Based on the selected decoding mode, the tool adjusts output timing determination, picture boundary detection, number of pictures in flight and/or jitter buffer utilization. For low-latency decoding, the tool can use a frame count syntax element to set initial output delay for a decoded picture buffer, and the tool can use auxiliary delimiter syntax elements to detect picture boundaries. To further reduce delay in low-latency decoding, the tool can reduce number of pictures in flight for multi-threaded decoding and reduce or remove jitter buffers. The tool receives encoded data, performs decoding according to the selected decoding mode to reconstruct pictures, and outputs the pictures for display.
摘要:
Techniques and tools described herein help manage memory efficiently during video decoding, especially when multiple video clips are concurrently decoded. For example, with clip-adaptive memory usage, a decoder determines first memory usage settings expected to be sufficient for decoding of a video clip. The decoder also determines second memory usage settings known to be sufficient for decoding of the clip. During decoding, memory usage is initially set according to the first settings. Memory usage is adaptively increased during decoding, subject to theoretical limits in the second settings. With adaptive early release of side information, the decoder can release side information memory for a picture earlier than the decoder releases image plane memory for the picture. The decoder can also adapt memory usage for decoded transform coefficients depending on whether the coefficients are for intra-coded blocks or inter-coded blocks, and also exploit the relative sparseness of non-zero coefficient values.
摘要:
A method and apparatus for scalable coding of a motion vector generated during motion estimation, in which a generated motion vector field is separated into a base layer and an enhancement layer according to pixel accuracies to obtain a layered structure for a motion vector. In addition, the motion vector field has a layered structure including a base layer composed of motion vectors of blocks larger than or equal to a predetermined size and at least one enhancement layer composed of motion vectors of blocks smaller than a predetermined size.
摘要:
A method and apparatus for mechanically splicing a pair of optic fibers or optic cables, the mechanical splice comprising: a ferrule having an axial capillary bore, the capillary bore configured to enclose the optic fibers at both ends of the ferrule; and cured epoxy disposed to secure together the ends of the optic fibers and to secure the optic fibers to an inside surface of the capillary bore, the ferrule optionally enclosed in a metal tube.
摘要:
Video decoding innovations for using local picture identifiers and computing co-located information are described. In one aspect, a decoder identifies reference pictures in a reference picture list of a temporal direct prediction mode macroblock that match reference pictures used by a co-located macroblock using local picture identifiers. In another aspect, a decoder determines whether reference pictures used by blocks are the same by comparing local picture identifiers during calculation of boundary strength. In yet another aspect, a decoder determines a picture type of a picture and based on the picture type selectively skips or simplifies computation of co-located information for use in reconstructing direct prediction mode macroblocks outside the picture.