Emmanuel Lancon

Pitch detection for use in a predictive speech coder

A pitch detector to adjust long term prediction in a pulse excitation speech coder. A residual signal r(n) is first derived from the speech signal s(n) by short term filtering. Then, r(n) is processed to calculate a prediction error signal e(n) which is subsequently pulse excitation encoded. The processing of e(n) entails prediction of a residual by measuring a pitch related factor M, employing two steps. First calculating a coarse M value through peak clipping and sign transition detection, and then adjusting the M value by autocorrelation--calculations about the roughly spaced peaks.

MPE/LTP speech coder for moblie radio application

Abstract In this paper, we describe our contribution to the CEPT effort for the definition of a European standard for the cellular mobile radio. We first give an overview of the Multipulse excitation with long term prediction (MPE/LTP) algorithm. Then, we report details on the implementation (constants, quantizing tables) respectively in the analysis part, the error protection/correction, and the synthesis part of the codec. Finally, we report detailed figures about the implementation efficiency (ROM/RAM/Cycles).

Low bit-rate hybrid coder using hierarchical motion compensation and low complexity vector quantization

An investigation is conducted of the problem of providing low-cost compression of video sequences to be used in applications like video sessions between workstations. An evaluation of a hybrid coder for very low bit-rate (64 Kb/s) coding, based on vector quantization of the motion-compensated prediction where only the regions that exhibit noticeable changes between successive frames are coded and transmitted, is presented. Motion compensation (MC) is performed in three steps, over successive subband decompositions of the subsequent frames. Such a hierarchical MC technique substantially reduces the complexity with respect to classical MC, due to the subsampling of the subbands. Vector quantization is applied to the coding of the motion-compensation prediction error blocks. Fast search VQ algorithms are used to achieve a low complexity coding.<<ETX>>

Sub-band coding of images using adaptive VQ and entropy coding

A study of a subband-based compression algorithm for still pictures is reported. A nonuniform band splitting is performed, then each subband signal is quantized by an adaptive vector quantizer with dynamic bit allocation. The output of each vector quantizer is processed by an arithmetic coder. Simulation results are reported, and it is shown that images of size 256*256 pixels with 265 gray levels can be encoded with excellent quality at rates between 0.5 and 1 bit/pixel and with good quality at rates below 0.3 bit pixel.<<ETX>>