Non-linear predictive vector quantization of speech
This is an incremental contribution to speech coding for audio processing applications.
The authors tackled speech coding by proposing a Non-Linear Predictive Vector Quantizer (PVQ) based on Multi-Layer Perceptrons and a method to evaluate quantizer design and correlation exploitation. The results showed no improvement over non-linear scalar predictors, but indicated potential for PVQ enhancement.
In this paper we propose a Non-Linear Predictive Vector quantizer (PVQ) for speech coding, based on Multi-Layer Perceptrons. We also propose a method to evaluate if a quantizer is well designed, and if it exploits the correlation between consecutive outputs. Although the results of the Non-linear PVQ do not improve the results of the non-linear scalar predictor, we check that there is some room for the PVQ improvement.