Lucent Technologies Inc.

At&t Bell Laboratories

At+t Corp.

American Telephone & Telegraph Co., At&t Bell Laboratories

American Telephone and Telegraph, Co.

At&t Ipm Corp.

Eg Technology, Inc.

John Hartung

Arnaud Eric Jacquin

Yi-Jen Chiu

Jonathan David Rosenberg

Donald Lars Duttweiler

Christine Irene Podilchuk

Thomas A Michel

Richard Dennis Gitlin

Robert J Safranek

Sam John

Nikil Jayant

Steven L Gay

David Malah

James Steven Pawlyk

Johathan D Rosenberg

Chris Elliot

James Pawlky

A method and apparatus for advantageously selecting video frames to be coded in order to improve the coding quality of a low bit-rate coder. In particular, temporal sub-sampling (i.e., selecting a set of frames to be coded from the complete incoming sequence of frames) is performed so that the frames which are to be coded are advantageously selected based upon a coding criterion, such as, for example, prediction gain (i.e., reduction in DFD variance). Specifically, in one illustrative embodiment, a larger number of frames are advantageously selected during periods of fast change, and correspondingly fewer frames are selected during other periods, while thereby keeping the overall apparent frame-rate fixed. The fixed frame-rate may, for example, be maintained by grouping the incoming sequence of frames into sequential groups of M consecutive frames, and then selecting exactly one frame per every M input frames, while permitting the selected frame to be at any advantageously selected location within the group of M frames. Thus, non-uniform frame selection is achieved, even though exactly one frame is actually coded within each superframe. Moreover, by basing the specific frame selection on an appropriate coding criterion (e.g., prediction gain), a substantial improvement in coder performance may be achieved for those critical portions of the video sequence during which a conventional coder's performance may be drastically reduced, without changing the apparent frame-rate.

Method and apparatus for video frame selection for improved coding quality at low bit-rates

The next generation digital signal processors and reduced instruction set chip processors have multiple arithmetic and logic units and multiply and accumulate units to boost the processor performance. This technique along with higher clock speed of the processors has moved the processor throughput bottle neck to the bus bandwidth. Thus, a bandwidth efficient implementation of standard processing functions is very important to improve the real throughput of these processors. A few processors already use a wide data bus to access multiple consecutive data elements in memory to boost the bus bandwidth. This feature has been used extensively to speed up the cache performance. Disclosed herein is a bandwidth efficient implementation of FIR filter. In any intensive signal processing applications, filters are the basic processing function. The proposed method reduces the required bus bandwidth by about a factor of 2.7 without increasing the number of MAC operations. This can significantly increase the processor real throughput in a RISC type of processor if it is used with an optimum processor architecture. This method can be extended to implement many type of filters, such as a complex filter, a convolution filter, a decimating filter and an adaptive filter.

Method and apparatus for performing rapid convolution

A method for estimating motion vectors between a current frame and a reference frame of a video signal wherein the current frame is divided into a plurality of search blocks and the reference frame and is defined to include a plurality of search regions corresponding to the plurality of search blocks, each of the plural search regions having a plurality of candidate blocks. The method includes the steps of calculating horizontal and vertical features of a reference frame based on pixel data of the reference frame; storing the calculated horizontal and vertical features; computing horizontal and vertical features of a k search block of the plurality of search blocks; comparing the horizontal and vertical features of the k search block with those of each of the plurality of candidate blocks of the reference frame in a corresponding search region using a computationally-efficient error measure; selecting a candidate block which yields a minimal error measure; and computing a motion vector for the k search block based on the selected candidate block.

Motion vector estimation based on statistical features of an image frame

A method of encoding a video sequence including a sequence of video images includes first comparing elements of a portion of a first video image (e.g., pixels of a macroblock of a current frame) with elements of a portion of a second video image (e.g., corresponding pixels of a macroblock of a previous frame) to generate respective intensity difference values for the element comparisons. Then, a first value (e.g., one) is assigned to the intensity difference values that are above a visually perceptible threshold value and a second value (e.g., zero) is assigned to the intensity difference values that are at or below the visually perceptible threshold value. Next, the method includes summing the first and second values to generate a sum. If the sum is greater than a predetermined decision value, the portion of the first video image is encoded (e.g., motion compensated). The method is fully compatible and, thus, may be implemented with video standards such as, for example, H.261, H.263, Motion-JPEG, MPEG-1, and MPEG-2.

Perceptual-based spatio-temporal segmentation for motion estimation

An encoder for coding multiple layers of video signal to provide for efficient use of bandwidth. The encoder produces an error signal based at least in part on criteria considered by a perceptual preprocessor which takes into account limitations inherent in human visual perception. If the error signal generated from a subject frame represents high priority data that would be visually perceptible by a viewer, as determined by the preprocessor, the error signal is coded onto a base layer and an error resilience layer of a coded signal for transmission. In the event that the error signal is determined to be of less priority, such as if it represents only slightly visually perceptible information, then the error signal is coded onto only a low priority layer of the coded signal.

Layered video coding using perceptual coding criteria for error resilience in packet networks

A method and apparatus for providing high-speed data compressing without sacrificing the quality of data reconstruction. Each input vector or block of original data is expressed as a combination of a codebook index and an error differential, or as a compressed version of the original block of data, depending on whether the total number of bits needed to express the input vector as a combination of a codebook index and an error differential is less than the total bits needed to send the compressed version of the original block of data. In one embodiment, the flexibility to express video data in a compressed format or as a combination of a codebook index plus an error differential is provided through a video system employing a codebook, a scalar quantizer, and an entropy coder.

Method and apparatus for improving vector quantization performance

A method and apparatus for coding an image comprising (at least) two regions (e.g., a facial region and a background region) which are to be coded with different quantization error parameters (e.g., quantizer step sizes), wherein the respective quantization error parameters are determined with use of a feed-forward rate control technique. Initial quantization error parameters are assigned to each region, and the overall bit rate which would result for the coding of the given frame is computed based on these assigned quantization error parameters. The computed bit rate is compared to a given bit rate constraint which may, for example, represent a limitation of a transmission channel. Based on the result of this comparison, one or both of the assigned quantization error parameters are adjusted, thereby resulting in different quantization error parameters for the two regions. For example, if the computed bit rate exceeds the given bit rate constraint, at least one of the quantization error parameters is advantageously adjusted in a direction which results in the use of fewer coding bits (e.g., a coarser quantization). If the given bit rate constraint exceeds the computed bit rate, at least one of the quantization error parameters is advantageously adjusted in a direction which results in the use of more coding bits (e.g., a finer quantization resulting in a higher coding quality). Finally, each of the regions is coded based on the correspondingly assigned quantization error parameters.

Global rate control for model-assisted coding of low bit rate video

A method and apparatus for performing vector quantization in a video coder. The method includes obtaining a video frame and segmenting it into coding blocks of predetermined size. A predetermined number of the coding blocks is transferred to a first cache memory. A predetermined portion of a codebook is transferred into a first section of a second cache memory which is capable of being rotated or having addresses associated with bytes in the second cache memory interchanged or offset by a specified amount. Concurrently with computing the vector quantization for the first portion of the codebook with respect to the coding blocks in the first cache memory, a second section of the codebook is transferred into a second section of the second cache memory. The number of coding blocks transferred to the first cache memory is selected so that transferral of the second section of the codebook is completed at the same time as or prior to completion of the computation of the vector quantization. The second cache memory may be rotated so that the process may be repeated for the remaining sections of the codebook with minimal or no delay, thereby permitting vector quantization to be performed in real time video coders.

Method and apparatus for vector quantization caching in a real time

A method of coding a set of multi- level signals, such as video signals comprises filtering the set of multi-level signals into one or more sets of multi-level sub-band signals; allocating a first quantity of bits for use in coding a first set of sub-band signals; coding the first set of multi-level sub-band signals using zero or more of the first quantity of allocated bits; allocating a second quantity of bits for use in coding a second set of sub-band signals, the second quantity of bits comprising zero or more of the first quantity of bits not used in coding the first set of multi-level sub-band signals; and coding the second set of multi-level sub-band signals using zero or more of the second quantity of allocated bits. Three examples of the method am provided. A hardware implementation involving the use of digital signal processors operating in parallel is also provided.

Method and apparatus for synthesizing subband video images

A method of coding a video signal at a fixed bit rate with use of a vector quantization codebook, wherein advantageous codebook updates are determined by comparing a coding error associated with a first provisional coding of a frame with a coding error associated with a second provisional coding of the frame. The first provisional coding uses a fixed available number of bits to code a frame with use of the unmodified codebook, whereas the second provisional coding uses a modified codebook to code the same frame. However, the second provisional coding codes both the frame and the codebook updates with the fixed available number of bits. The provisional coding having the smaller coding error is selected as the coding of the frame which will be adopted. In addition, the codebook may be replaced with the modified codebook when the second provisional coding is selected. In this manner, the coder will adapt the codebook to the characteristics of the scene.

Adaptive video coder with dynamic bit allocation

The present invention provides a method of coding a set of multi-level signals, such as video signals. The method comprises filtering the set of multi-level signals into one or more sets of multi-level sub-band signals; allocating a first quantity of bits for use in coding a first set of sub-band signals; coding the first set of multi-level sub-band signals using zero or more of the first quantity of allocated bits; allocating a second quantity of bits for use in coding a second set of sub-band signals, the second quantity of bits comprising zero or more of the first quantity of bits not used in coding the first set of multi-level sub-band signals; and coding the second set of multi-level sub-band signals using zero or more of the second quantity of allocated bits. Three examples of the method are provided. A hardware implementation involving the use of digital signal processors operating in parallel is also provided.

Dynamic bit allocation for three-dimensional subband video coding

In a private telecommunications network, a plurality of digital PBXs are interconnected via pairs of codecs. The codecs of each pair, or 'tandem', are each operative to encode 64 kilobit/second (kbps) mu-law speech so as to compress it to 16 kbps speech for transmission to the other codec in the tandem. The latter is operative to thereafter decompress the 16 kbps speech back to 64 kbps. Each codec has a second mode of operation in which, rather than decode the encoded speech, it preserves the bits thereof in its own output signal. The codec transitions to this mode whenever it recognizes the presence of another codec on its high-bit-rate side of the connection. As a result, only one encoding/decoding cycle is performed across the connection, thereby minimizing the speech-coding-induced distortion and delay therein. The mechanism enabling a codec to communicate its presence to another codec on its high-bit-rate side of the connection is based on the transmission of predetermined synchronization patterns inserted in the signals it outputs in that direction. In a second embodiment, codecs of the above-described type are used in a cellular mobile radio telecommunications system.

Codecs with suppression of multiple encoding/decodings across a

Impulse response estimation is improved in an adaptive filter by compensating digital signals incoming to the filter to include a nonlinearity substantially identical to a nonlinearity included in a return signal path to the filter.

Adaptive filter including signal path compensation

Update gain normalization is employed in adaptive filters to control weighting of the filter impulse response updates in order to converge properly to a desired impulse response. Singing, i.e., oscillating, of the filter is overcome by adjusting the update gain when an incoming signal power estimate used to normalize the gain exceeds a prescribed threshold value. In one example, the normalized gain is adjusted to be a fixed value for power estimate values which exceed the threshold. In accordance with another aspect of the invention, a single normalized gain value is used to adjust the update gain in two adaptive filters employed as echo cancelers in a bidirectional voice frequency repeater.

Growing community of inventors

Warren, NJ, United States of America

John Hartung