An adaptive differential pulse-code-modulation speech coder with a switched predictor adaptation scheme is described. The predictor adaptation is done by switching one of the several predetermined predictors stored bo...
详细信息
An adaptive differential pulse-code-modulation speech coder with a switched predictor adaptation scheme is described. The predictor adaptation is done by switching one of the several predetermined predictors stored both at the encoder and the decoder. The selection of the predictor is controlled by a vector quantiser. The tradeoff between the performance and the extra side information that needs to be transmitted is evaluated for bit rates of 16, 24 and 32 Kbit/s.
Although a monochromatic plane wave of a certain circular polarization state and wavelength may be almost completely reflected by a chiral sculptured thin film, we show here that information coded in rectangular pulse...
详细信息
Although a monochromatic plane wave of a certain circular polarization state and wavelength may be almost completely reflected by a chiral sculptured thin film, we show here that information coded in rectangular pulses can still he transmitted across the film-provided the rise and fall rimes of the pulses are sufficiently small and the pulse widths are sufficiently large. (C) 2001 John Wiley & Sons, Inc.
Gray-coding of signal amplitudes can protect PCM transmission of analogue signals from the effects of channel bit errors. We generalise this to VQ, and provide a `vector switching?? algorithm that reduces the expected...
详细信息
Gray-coding of signal amplitudes can protect PCM transmission of analogue signals from the effects of channel bit errors. We generalise this to VQ, and provide a `vector switching?? algorithm that reduces the expected signal distortion by suitably assigning code words to the codevectors in a given code-book.
An adaptive differential pulsecodemodulation (ADPCM) system is described having a set of fixed vector predictors, and a quantiser that is adapted in the forward direction with the distinction that its step size is p...
详细信息
An adaptive differential pulsecodemodulation (ADPCM) system is described having a set of fixed vector predictors, and a quantiser that is adapted in the forward direction with the distinction that its step size is predicted from previous values. SNR gains compared to DPCM-AQF are presented for bit rates of 16, 24 and 32 kbit/s.
The sources of bit-rate variation in a conditional-replenishment videoconference coder were found experimentally to be due to the interaction between coding method and image content. The isolation of the components of...
详细信息
The sources of bit-rate variation in a conditional-replenishment videoconference coder were found experimentally to be due to the interaction between coding method and image content. The isolation of the components of variation and the effect on them of buffering and packetisation are reported.
A very high-quality sub-band ADPCM music coding scheme which compresses high-fidelity music signals, bandlimited to 15 kHz, to an equivalent PCM representation of only 4 bits per sample, is described. By processing mu...
详细信息
A very high-quality sub-band ADPCM music coding scheme which compresses high-fidelity music signals, bandlimited to 15 kHz, to an equivalent PCM representation of only 4 bits per sample, is described. By processing music sampled at 32 kHz, this coder exhibits a total bit rate of only 128 kbit/s and is consequently applicable to the ISDN. Subjective tests conducted with this coder have shown that music recovered from the compression scheme is essentially indistinguishable from the original material. The results obtained are of major importance, not only for ISDN and broadcasting, but also for other digital audio technology such as compact disc (CD) and digital audio tape
A technique for nonstandard A/D conversion of signals subject to a fidelity criterion is described, based on Hopfield neural networks. Switched-capacitor neural networks offer a natural means for efficient implementat...
详细信息
A technique for nonstandard A/D conversion of signals subject to a fidelity criterion is described, based on Hopfield neural networks. Switched-capacitor neural networks offer a natural means for efficient implementation of the proposed technique. Applications include all forms of PCM coding, oversampled A/D conversion and digital image halftoning
A method for the design of cost-effective one-stage PCM switching networks composed of uniform time-space elements is proposed. These one-stage networks can be used as submodules for multistage networks, having, in ma...
详细信息
A method for the design of cost-effective one-stage PCM switching networks composed of uniform time-space elements is proposed. These one-stage networks can be used as submodules for multistage networks, having, in many cases, fewer elements than currently known networks.
A systematic way of designing efficient trellis coded modulation (TCM) is presented. Design is based on the concept of trellis partitioning and newly defined branch distances. A lower bound on free Euclidean distance ...
详细信息
A systematic way of designing efficient trellis coded modulation (TCM) is presented. Design is based on the concept of trellis partitioning and newly defined branch distances. A lower bound on free Euclidean distance is given.
Presents a fast DPCM coder built around a ROM of feasible size, in which an elementary form of tree encoding is integrated, to improve the quality of the decoded picture. Simulations illustrate the improvements of thi...
详细信息
Presents a fast DPCM coder built around a ROM of feasible size, in which an elementary form of tree encoding is integrated, to improve the quality of the decoded picture. Simulations illustrate the improvements of this technique in comparison with normal DPCM encoding
暂无评论