Enhancement Of Coded Speech Using a Mask-Based Post-Filter

TL;DR

This paper introduces a neural network-based post-filter that enhances low-bitrate coded speech by estimating time-frequency masks, outperforming traditional heuristic methods in both objective and subjective evaluations.

Contribution

It proposes a data-driven, mask-based post-filter using neural networks for speech enhancement at low bitrates, demonstrating superior performance over conventional filters.

Findings

Neural network post-filters improve speech quality at low bitrates.

Mask-based neural models outperform heuristic post-filters.

Objective and subjective tests confirm the effectiveness of the proposed method.

Abstract

The quality of speech codecs deteriorates at low bitrates due to high quantization noise. A post-filter is generally employed to enhance the quality of the coded speech. In this paper, a data-driven post-filter relying on masking in the time-frequency domain is proposed. A fully connected neural network (FCNN), a convolutional encoder-decoder (CED) network and a long short-term memory (LSTM) network are implemeted to estimate a real-valued mask per time-frequency bin. The proposed models were tested on the five lowest operating modes (6.65 kbps-15.85 kbps) of the Adaptive Multi-Rate Wideband codec (AMR-WB). Both objective and subjective evaluations confirm the enhancement of the coded speech and also show the superiority of the mask-based neural network system over a conventional heuristic post-filter used in the standard like ITU-T G.718.