Learning More with Less: Self-Supervised Approaches for Low-Resource Speech Emotion Recognition
This work addresses the challenge of limited annotated data for speech emotion recognition in underrepresented languages, with incremental improvements in performance.
The paper tackled the problem of speech emotion recognition for low-resource languages by exploring self-supervised learning approaches, achieving F1 score improvements of 10.6% in Urdu, 15.2% in German, and 13.9% in Bangla.
Speech Emotion Recognition (SER) has seen significant progress with deep learning, yet remains challenging for Low-Resource Languages (LRLs) due to the scarcity of annotated data. In this work, we explore unsupervised learning to improve SER in low-resource settings. Specifically, we investigate contrastive learning (CL) and Bootstrap Your Own Latent (BYOL) as self-supervised approaches to enhance cross-lingual generalization. Our methods achieve notable F1 score improvements of 10.6% in Urdu, 15.2% in German, and 13.9% in Bangla, demonstrating their effectiveness in LRLs. Additionally, we analyze model behavior to provide insights on key factors influencing performance across languages, and also highlighting challenges in low-resource SER. This work provides a foundation for developing more inclusive, explainable, and robust emotion recognition systems for underrepresented languages.