Abstract:Recent years have witnessed the exponential increase in the demand for ultra high definition (UHD) video compression. The third generation of Audio Video Coding Standard (AVS3), which is also known as IEEE Standard 1857.10, is the latest audio and video coding standard developed by the China AVS working group. In AVS3, targeting for UHD videos, a series of efficient coding tools have been introduced, leading to the dramatical increase of computational burden. In this scenario, real-time decoding of UHD videos becomes extremely challenging. This paper presents an improved hybrid CPU + GPU accelerated framework for AVS3 decoding. In particular, the motion vector (MV) derivation process is extracted from entropy decoding threads on the CPU. Therefore, the dependency between threads is removed and the entropy decoding can be performed by multiple threads efficiently. Regarding the GPU, we design compact data structures for transform, prediction, and in-loop filtering to reduce the burden of data transmission. A flexible information buffer supporting multi-thread random writing is further created to coordinate the computation between CPU and GPU. Through asynchronous operations on the buffer, the computation of different computing units and the data transmission between them could be performed in parallel. With NVIDIA GeForce RTX 2080Ti GPU and Intel Core i7 8700K CPU, the proposed decoder achieves 151 frames per second (fps) for 4K videos and 55 fps for 8K videos in all intra configuration. In random access configuration, 218 fps and 74 fps are obtained for 4K and 8K videos, respectively.

A 2Gpixel/s H.264/AVC HP/MVC video decoder chip for Super Hi-Vision and 3DTV/FTV applications

An Efficient Hardware Design for HDTV H.264/AVC Encoder

Parallel Enhanced Low Design Effort H.264/AVC Fractional Motion Estimation Engine for Super Hi-Vision Application

Introduction to the Applications of H.264/AVC Decoder Chips

Pipelined Architecture Design of H.264/AVC CABAC Real-Time Decoding

An AVS HDTV Video Decoder Architecture Employing Efficient HW/SW Partitioning.

A real-time multi-view interlacing architecture for auto-stereoscopic 3DTV display based on FPGA

A Real-Time Full HD 2D-to-3d Video Conversion System Based on FPGA

A High-Performance 2-D Inverse Integer Transform Architecture for AVS

VLSI Architecture for H.264 Integer-Pel Motion Estimation with Minimum Memory Bandwidth Requirement

High Definition Real-Time Decoding

A High-Performance VLSI Architecture for CABAC Decoding in H.264/AVC

Full High-Definition Real-Time Depth Estimation for Three-Dimensional Video System

A Single-Chip 4K 60-fps 4:2:2 HEVC Video Encoder LSI Employing Efficient Motion Estimation and Mode Decision Framework With Scalability to 8K

Hardware Solution of Real-Time Depth Estimation Based on Stereo Vision

High Throughput and Low Memory Access Sub-Pixel Interpolation Architecture for H.264/AVC HDTV Decoder

Full Hd Real-Time Depth Estimation Algorithm and Hardware Implementation

High Performance Multi Transform Coding Hardware Architecture Design for H.264/AVC

A Real-Time AVS3 8K-UHD Encoding and Decoding System

A 4K $\times$ 2k@60fps Multifunctional Video Display Processor for High Perceptual Image Quality

A Real-Time Ultra-High Definition Video Decoder of AVS3 on Heterogeneous Systems