Correlation Distance Skip Connection Denoising Autoencoder (CDSK-DAE) for Speech Feature Enhancement

TL;DR

This paper introduces a novel correlation distance skip connection denoising autoencoder (CDSK-DAE) for speech feature enhancement, significantly improving noise robustness in end-to-end ASR systems across various noisy environments.

Contribution

It proposes a new autoencoder architecture with skip connections and a correlation distance-based training objective, enhancing speech features for noise-robust ASR.

Findings

Outperforms conventional and state-of-the-art models in noisy conditions

Improves word error rate (WER) across multiple noise types and SNR levels

Effective with both linear and non-linear penalty terms

Abstract

Performance of learning based Automatic Speech Recognition (ASR) is susceptible to noise, especially when it is introduced in the testing data while not presented in the training data. This work focuses on a feature enhancement for noise robust end-to-end ASR system by introducing a novel variant of denoising autoencoder (DAE). The proposed method uses skip connections in both encoder and decoder sides by passing speech information of the target frame from input to the model. It also uses a new objective function in training model that uses a correlation distance measure in penalty terms by measuring dependency of the latent target features and the model (latent features and enhanced features obtained from the DAE). Performance of the proposed method was compared against a conventional model and a state of the art model under both seen and unseen noisy environments of 7 different types of background noise with different SNR levels (0, 5, 10 and 20 dB). The proposed method also is tested using linear and non-linear penalty terms as well, where, they both show an improvement on the overall average WER under noisy conditions both seen and unseen in comparison to the state-of-the-art model.