A5056088829- Video Coding and Compression Technologies
- Advanced Vision and Imaging
- Advanced Data Compression Techniques
- Image and Video Quality Assessment
- Video Analysis and Summarization
- Advanced Image Processing Techniques
- Image and Signal Denoising Methods
- Quantum Information and Cryptography
- Quantum optics and atomic interactions
- Cold Atom Physics and Bose-Einstein Condensates
- Image and Video Stabilization
- Digital Filter Design and Implementation
- Telecommunications and Broadcasting Technologies
- Image Enhancement Techniques
- Manufacturing Process and Optimization
- Data Management and Algorithms
- VLSI and FPGA Design Techniques
- Catalysis and Oxidation Reactions
- Embedded Systems Design Techniques
- Computer Graphics and Visualization Techniques
- Zeolite Catalysis and Synthesis
- Multimedia Communication and Technology
- Advanced NMR Techniques and Applications
- Spectroscopy and Chemometric Analyses
- Quantum and electron transport phenomena
MediaTek (Taiwan)
2007-2024
National Tsing Hua University
2006-2022
MediaTek (China)
2011-2013
National Taiwan University
2003-2008
University of Southern California
2006
This paper provides a technical overview of newly added in-loop filtering technique, sample adaptive offset (SAO), in High Efficiency Video Coding (HEVC). The key idea SAO is to reduce distortion by first classifying reconstructed samples into different categories, obtaining an for each category, and then adding the category. category properly calculated at encoder explicitly signaled decoder reducing effectively, while classification performed both saving side information significantly. To...
H.264/AVC significantly outperforms previous video coding standards with many new tools. However, the better performance comes at price of extraordinarily huge computational complexity and memory access requirement, which makes it difficult to design a hardwired encoder for real-time applications. In addition, due complex, sequential, highly data-dependent characteristics essential algorithms in H.264/AVC, both pipelining parallel processing techniques are constrained be employed. The...
Variable block-size motion estimation (VBSME) has become an important video coding technique, but it increases the difficulty of hardware design. In this paper, we use inter-/intra-level classification and various data flows to analyze impact supporting VBSME in different architectures. Furthermore, propose two architectures that can support traditional fixed as well with less chip area overhead compared previous approaches. By broadcasting reference pixel rows propagating partial sums...
Adaptive loop filtering for video coding is to minimize the mean square error between original samples and decoded by using Wiener-based adaptive filter. The proposed ALF located at last processing stage each picture can be regarded as a tool catch fix artifacts from previous stages. suitable filter coefficients are determined encoder explicitly signaled decoder. In order achieve better efficiency, especially high resolution videos, local adaptation used luma signals applying different...
This paper presents an overview of the technologies for in-loop processing and filtering in Versatile Video Coding (VVC) standard. These processes comprise luma mapping with chroma scaling, deblocking filter, sample adaptive offset, loop filter cross-component filter. They are qualified as "in-loop" because they applied inside encoding decoding loops, before storing pictures decoded picture buffer. The filters complementary address different purposes. Luma scaling aims at adaptively...
An H.264/AVC encoder is implemented on a 31.72mm/sup 2/ die with 0.18/spl mu/m CMOS technology. A four-stage macroblock pipelined architecture encodes 720p 30f/s HDTV videos in real time at 108MHz. The encoded video quality competitive reference software requiring 3.6TOPS general-purpose processor-based platform.
The memory bandwidth reduction for motion estimation is important because of the power consumption and limited in video coding systems. In this paper, we propose a Level C+ scheme which can fully reuse overlapped searching region horizontal direction partially vertical to save more compared C scheme. However, direct implementation may conflict with some tools then induces lower hardware efficiency Therefore, n-stitched zigzag scan discuss two types 2-stitched MPEG-4 H.264 as examples. They...
A new video coding tool, sample adaptive offset (SAO), is introduced in this paper. SAO has been adopted into the Working Draft of standard, High-Efficiency Video Coding (HEVC). The located after deblocking loop. concept to classify reconstructed pixels different categories and then reduce distortion by simply adding an for each category pixels. pixel intensity edge properties are used classification. To further improve efficiency, a picture can be divided regions localization parameters....
This paper describes a video coding technology proposal submitted by MediaTek in April 2018 response to joint call for proposals (CfP) issued ITU-T VCEG and ISO/IEC MPEG. The uses the conventional block-based hybrid approach with breakthrough design of block structure, substantial improvements tools, new inventions loop filtering techniques. First, structure is designed using highly flexible partitioning scheme, where each tree unit picture split into multiple units (CUs) recursive...
A de-interlacing algorithm using adaptive 4-field global/local motion compensated approach is presented. It consists of block-based directional edge interpolation, same-parity detection, estimation and compensation. The edges are sharper when the interpolation adopted. same parity detection local detect static areas fast by four reference fields, global detects camera panning zooming motions. compensation recover interlaced videos to progressive ones. Experimental results show that peak...
Phosphorus modification of a HZSM-5 (MFI) zeolite by wet impregnation has long been known to decrease aromatic formation in methanol conversion chemistry. We prepared and studied catalyst modified introducing trimethylphosphine under reaction conditions followed oxidation. Magic-angle spinning (MAS) NMR shows that extensive dealumination occurs, resulting with much higher framework SiO2/Al2O3 ratio, as well extraframework aluminum approximately 1.4 equiv entrained phosphoric acid (under...
VLSI realization of video compression is the key to many real-time multimedia communications systems. Among algorithms, newly established MPEG-4 and, in particular, H.264 standards have become increasingly popular. However, high coding efficiency such algorithms comes at cost a dramatic increase complexity. Effective and efficient hardware solutions this problem are necessary. In article we present an overview design issues H.264. Both module system architectures these two discussed. Based...
A computation-aware motion estimation algorithm is proposed in this paper. Its goal to find the best block-matching results a computation-limited and computation-variant environment. Our characterized by one-pass flow with adaptive search strategy. In prior scheme, Tsai et al. propose that all macroblocks are processed simultaneously, more computation allocated macroblock largest distortion among entire frame step-by-step fashion. This implies random access of required, related information...
The on-chip line buffer dominates the total area and power of line-based 2-D discrete wavelet transform (DWT). In this paper, a memory-efficient VLSI implementation scheme for DWT is proposed, which consists two parts, wordlength analysis methodology multiple-lifting scheme. required memory determined firstly by use proposed methodology, DWT, named scheme, then proposed. can guarantee to avoid overflow coefficients, average difference between predicted experimental quality level only 0.1 dB...
This article introduces adaptive loop filtering (ALF) techniques being considered for the HEVC standard. The key idea of ALF is to minimize mean square error between original pixels and decoded using Wiener-based filter coefficients. located at last processing stage each picture can be regarded as a tool trying catch fix artifacts from previous stages. suitable coefficients are determined by encoder explicitly signaled decoder. In order achieve better coding efficiency, especially high...
Loading quantum information deterministically onto a node is an important step toward network. Here, we demonstrate that coherent-state microwave photons with optimal temporal waveform can be efficiently loaded single superconducting artificial atom in semi-infinite one-dimensional (1D) transmission-line waveguide. Using weak coherent state (the number of (N) contained the pulse ≪1) exponentially rising waveform, whose time constant matches decoherence atom, loading efficiency 94.2% ± 0.7%...
In this paper, a one-pass encoding algorithm is proposed for adaptive loop filter (ALF) in high-efficiency video coding (HEVC). ALF can improve both subjective and objective quality, but it also requires lot of passes (i.e. picture buffer accesses) that will significantly increase external memory access, latency, power consumption. Therefore, we propose method to estimate filtering distortion without performing real operation. The number be effectively reduced from 16 1. Combined with an...
To the best of authors' knowledge, this paper presents first work on memory analysis VLSI architectures for motion-compensated temporal filtering (MCTF). The open-loop MCTF prediction scheme has led revolution hybrid video coding methods that are mainly based close-loop MC (MCP) scheme, and it also becomes core technology coming standard, MPEG-21 part 13-scalable (SVC). In paper, macroblock (MB)-level frame-level data reuse schemes analyzed MCTF. MB-level is especially motion estimation...
In this article, we suggest some techniques to design the H.264/AVC video coding system for HDTV applications. The exploration is made according software profiling. considerations of scheduling and pipelining are discussed followed by architecture optimization significant modules. efficient achieved combining these
Since motion-compensated temporal filtering (MCTF) becomes an important prediction scheme in video coding algorithms, this paper presents efficient engine which not only is the first MCTF hardware work but also supports traditional (MCP) to provide computation scalability. For stage of and MCP schemes, modified extended double current Frames adopted reduce system memory bandwidth, a frame-interleaved macroblock pipelining proposed eliminate induced data buffer overhead. In addition, update...
A motion compensated de-interlacing method with adaptive global estimation and compensation is proposed to recover the defects of interlaced video sequence camera panning, rotating or zooming. GME GMC are used change whole picture due motions. Two local methods applied de-interlace sequences without respectively. SAD checking global/local comparator can be as a block based mode decision system for intra/MC/GMC modes. The algorithm could achieve higher image quality than any other usual on...
A computation-aware motion estimation algorithm is proposed. Its goal to find the best block matching results in a computation-limited and computation-variant environment. Our new features are one-pass flow adaptive search strategies. The prior scheme allocates more computation macroblock with highest distortion entire frame step by step. This implies that random access of macroblocks inevitable, pattern must be determined advance. requires huge size memory for all store up-to-date minimum...