TRNet: Two-level Refinement Network leveraging Speech Enhancement for Noise Robust Speech Emotion Recognition

TL;DR

TRNet is a novel two-level refinement network that uses speech enhancement techniques to improve noise robustness in speech emotion recognition, maintaining performance across noisy and noise-free conditions.

Contribution

Introduces TRNet, a two-level refinement approach combining speech enhancement and deep representation refinement for robust SER in noisy environments.

Findings

Significantly improves SER accuracy in noisy conditions

Maintains performance in noise-free environments

Effective in both matched and unmatched noise scenarios

Abstract

One persistent challenge in Speech Emotion Recognition (SER) is the ubiquitous environmental noise, which frequently results in deteriorating SER performance in practice. In this paper, we introduce a Two-level Refinement Network, dubbed TRNet, to address this challenge. Specifically, a pre-trained speech enhancement module is employed for front-end noise reduction and noise level estimation. Later, we utilize clean speech spectrograms and their corresponding deep representations as reference signals to refine the spectrogram distortion and representation shift of enhanced speech during model training. Experimental results validate that the proposed TRNet substantially promotes the robustness of the proposed system in both matched and unmatched noisy environments, without compromising its performance in noise-free environments.