Sarnoff Corporation

National Yang Ming Chiao Tung University

Industrial Technology Research Institute

Sharp Corporation

Mediatek Corporation

Mitsubishi Electric Research Laboratories, Inc.

Thomson Licensing S.a.

National Cheng-kung University

Ambarella, Inc.

Thomson Multimedia S.a.

Ambarella Taiwan Ltd.

Tihao Chiang

Ya-Qin Zhang

Ravi Krishnamurthy

Chung-Neng Wang

Huifang Sun

Hung-Ju Lee

Shih-Hao Wang

Xudong Song

Sriram Sethuraman

Leslie D Kohn

Joel Walter Zdepski

Wilson Kwok

Xiaobing Lee

Max Chien

Tzu-Liang Su

Ming-Yen Huang

Anthony Vetro

Iraj Sodagar

Jungwoo Lee

Zixiang Xiong

Hsueh-Ming Hang

Hsiang-Chun Huang

Shin-Wei Yang

Chih-Chieh Chen

Chi-Min Liu

Jeng-Hung Luo

Chih-Hung Li

Gen-Ming Lee

Wen-Hsiao Peng

Sassan Pejhan

Paul Hatrack

Yao-Chung Lin

Zudong Song

A method for video encoding is disclosed. The method generally includes the steps of (A) generating first sub-pel data for at least one of (i) a motion estimation and (ii) a mode decision by first filtering reference data and (B) generating second sub-pel data for a motion compensation by second filtering the reference data. Wherein a first performance of the first filtering may be different than a second performance of the second filtering.

Mode decision using approximate &#xbd; pel interpolation

Mode decision using approximate 1/2 pel interpolation

An MPEG-4 system with error concealment is provided for video service under the network with packet loss. The MPEG-4 system includes an encoder and a decoder. The encoder uses an intra-refreshment technique is used to make coded bitstream more robust against noise in order to stop error propagation. The rate-distortion optimization criterion is also introduced to adaptively update in synchronization with intra-coded blocks adaptively based on the true network condition with minimal overhead. The Lagrange multiplier is modified to achieve the best rate-distortion balance. In addition, a decoder loop is used in the encoder and is synchronized with the true decoder to achieve the best performance and avoid mismatch with the decoder used in the MPEG-4 system. The decoder is able to achieve resilient decoding from any kind of noise and enhance the reconstructed image quality with spatial and temporal hybrid concealment method. The result shows that a 3.65-9.71 dB further improvement on peak-signal-to-noise-ratio (PSNR) can be achieved in comparison with the existing methods that adopt spatial copy and zero motion concealment in decoding.

MPEG-4 streaming system with adaptive error concealment

A prediction module includes a computation control unit and first to sixth processing units. The computation control unit arranges the pixels of a reference block outputted by a frame buffer appropriately according to data decoded by an entropy decoder into first and second pixel signals, and outputs the same to the first to sixth processing units in units of cycles. The processing units are controlled by the computation control unit to respectively complete corresponding computations in a cycle, and to use the computation results as values of first to sixth output signals to be correspondingly outputted in a next cycle. The computation control unit receives the first to sixth output signals, and computes the output signals so as to obtain pixel values of a macroblock to be predicted.

Prediction module

A method transcodes a compressed multi-layer video bitstream that includes a base layer bitstream and an enhancement layer bitstream. The base and enhancement layers are first partially decoded, and then the partially decoded signals are combined with a motion compensated signal yielding a combined signal. The combined signal is quantized into an output signal according to a quantization parameter, and the output signal is variable length encoded as a single layer bitstream. In a preprocessing step, the enhancement layer can be truncated according to rate control constraint, and the same constraints can also be used during the quantization.

Video transcoding of scalable multi-layer videos to single layer video

A method and apparatus for selecting a quantizer scale for each object within a frame to optimize the coding rate is disclosed. A quantizer scale is selected for each region or 'object' within each frame such that the target bit rate for the frame is achieved while maintaining a uniform visual quality over the entire frame.

Apparatus and method for object based rate control in a coding system

A robust fine granularity scalability video encoding includes a base layer encoder and an enhancement layer encoder in which motion compensated difference images are generated by comparing an original image to predicted images at base layer and enhancement layer with motion compensation. Based on leaky and partial predictions, a high quality reference image is constructed at the enhancement layer to improve temporal prediction. In the construction of the high quality reference image, one parameter &#x3b2; controls the number of bitplanes of the enhancement layer difference coefficients used and another parameter &#x3b1; controls the amount of predictive leak. A spatial scalability module allows the processed pictures at the base layer and the enhancement layer to have identical or different spatial resolutions.

Architecture and method for fine granularity scalable video coding

The present invention proposes a fast motion estimation using N-queen pixel decimation, whereby after a reference block and a block to be processed are selected in a video sequence, an N&#xd7;N queens pattern is used for pixel decimation to perform block match, thereby obtaining a good enough block difference value. The present invention combines pixel decimation with fast motion estimation for search points reduction to achieve the object of simplifying computational complexity of motion estimation. Therefore, the present invention can sieve out sufficiently representative pixels and will not increase extra computational complexity.

Fast motion estimation using N-queen pixel decimation

A method of motion estimation for video encoding constructs a binary pyramid structure having three binary layers. A state update module registers and updates repeat occurrence of final motion vectors and a static-state checking module determines if the method is in a static mode or a normal mode based on the repeat occurrence. In a normal mode, the first binary layer is searched within a &#xb1;3 pixel refinement window to determine a first level motion vector. In the second binary layer, a search range is computed based on six motion vector candidates. By checking every point within in the search range, a second binary layer search generates a second level motion vector. Finally, a third binary layer search within a &#xb1;2 pixel refinement window generates a final motion vector according to the second level motion vector. In a static mode, a fine tuning module performs search within a &#xb1;1 pixel refinement window and generates a final motion vector.

Method and apparatus for motion estimation with all binary representation

An apparatus and method for dynamically varying the frame rate of an image sequence is disclosed. In one embodiment, the image sequence is encoded and stored at different frame rates (e.g., 30, 25, 20 fps and so on). Alternatively, only the motion information, e.g., motion vectors, for the other frame rates are stored.

Apparatus and method for dynamically controlling the frame rate of video streams

A PC-type computer has a system bus (e.g., a PCI bus) configured with a main CPU board, a statistical multiplexing (stat-mux) board, and a plurality of video/audio encoder boards, each configured to receive and compress a corresponding video/audio stream. The stat-mux board performs statistical multiplexing on the different compressed bitstreams to transmit multiple bitstreams over individual shared communication channels. Although each of the boards is configured to the system bus, each encoder board has a digital signal processor (DSP) with a synchronized serial interface (SSI) output port that is directly connected to an SSI input port on a DSP on the stat-mux board (which, in one embodiment, has four such DSPs each with six such SSI input ports). As such, (up to 24) compressed video/audio bitstreams generated on the various encoder boards can be transmitted directly to the stat-mux board without having to go through the system bus. In this way, the computer system can provide statistical multiplexing of low-latency video/audio bitstreams without having to suffer the processing delays associated with conventional transmission over PCI system buses.

Computer system for statistical multiplexing of bitstreams

A method and apparatus for selecting a quantizer scale for each macroblock within a frame to optimize the coding rate is presented. A quantizer scale is selected for each macroblock within each frame such that the target bit rate for the frame is achieved while maintaining a uniform visual quality over the entire frame.

Apparatus and method for macroblock based rate control in a coding system

When two or more different video streams a e compressed for concurrent transmission of multiple compressed video bitstreams over a single shared communication channel, control over both (1) the transmission of data over the shared channel and (2) the compression processing that generates the bitstreams is exercised taking into account the differing levels of latency required for the corresponding video applications. For example, interactive video games typically require lower latency than other video applications such as video streaming, web browsing, and electronic mail. A multiplexer and traffic controller takes these differing latency requirements, along with bandwidth and image fidelity requirements, into account when controlling both traffic flow and compression processing. In addition, an off-line profiling tool analyzes typical video applications off-line in order to generate profiles of different types of video applications that are then accessed in real-time by a call admission manager responsible to controlling the admission of new video application sessions as well as the assignment of admitted applications to specific available video encoders, which themselves may differ in video compression processing power as well as in the degree to which they allow external processors (like the multiplexer and traffic controller) to control their internal compression processing.

Latency-based statistical multiplexing

An apparatus and a concomitant method for using M-ary pyramid decomposition in combination with N-scale tiling to reduce the computational complexity in determining motion vectors for block-based motion estimation is disclosed.

Apparatus and method for employing M-ary pyramids with N-scale tiling

Growing community of inventors

Taipei, Taiwan

Tihao Chiang