A comparative study of several ADPCM schemes with linear and nonlinear prediction

TL;DR

This paper compares various ADPCM schemes, including neural network-based nonlinear prediction and classical linear prediction, focusing on adaptive quantization methods and their performance differences.

Contribution

It provides a comparative analysis of ADPCM schemes with nonlinear neural network prediction versus traditional linear prediction, highlighting their relative advantages.

Findings

Neural network-based ADPCM schemes show improved prediction accuracy.

Adaptive quantization significantly impacts compression efficiency.

Comparison reveals specific scenarios where nonlinear prediction outperforms linear methods.

Abstract

In this paper we compare several ADPCM schemes with nonlinear prediction based on neural nets with the classical ADPCM schemes based on several linear prediction schemes. Main studied variations of the ADPCM scheme with adaptive quantization (2 to 5 bits) are: -forward vs backward -sample adaptive vs block adaptive