Feature Selection Enhancement and Feature Space Visualization for Speech-Based Emotion Recognition

TL;DR

This paper introduces a feature enhancement and visualization method for speech emotion recognition that improves accuracy by applying PCA and t-SNE to selected feature subsets, validated on two multilingual datasets.

Contribution

The study presents a novel feature enhancement strategy combining PCA and feature fusion, along with visualization, to improve speech emotion recognition accuracy.

Findings

Achieved 11.5% average recognition gain on EMO-DB dataset.

Achieved 13.8% average recognition gain on RAVDESS dataset.

Validated effectiveness across German and English emotional speech datasets.

Abstract

Robust speech emotion recognition relies on the quality of the speech features. We present speech features enhancement strategy that improves speech emotion recognition. We used the INTERSPEECH 2010 challenge feature-set. We identified subsets from the features set and applied Principle Component Analysis to the subsets. Finally, the features are fused horizontally. The resulting feature set is analyzed using t-distributed neighbour embeddings (t-SNE) before the application of features for emotion recognition. The method is compared with the state-of-the-art methods used in the literature. The empirical evidence is drawn using two well-known datasets: Emotional Speech Dataset (EMO-DB) and Ryerson Audio-Visual Database of Emotional Speech and Song (RAVDESS) for two languages, German and English, respectively. Our method achieved an average recognition gain of 11.5\% for six out of seven emotions for the EMO-DB dataset, and 13.8\% for seven out of eight emotions for the RAVDESS dataset as compared to the baseline study.