MERaLiON-SER: Robust Speech Emotion Recognition Model for English and SEA Languages
/ Authors
Hardik B. Sailor, AiTi Aw, Nancy F. Chen, Ying Lay Chiu, Yang Ding, Yingxu He, Ridong Jiang, Jingtao Li, Jingyi Liao, Zhuohan Liu
and 19 more authors
Yanfeng Lu, Yi Ma, Manas Gupta, Muhammad Huzaifah bin Md Shahrin, Nabilah Binte Md Johan, Nattadaporn Lertcheva, Chunlei Pan, Minh Duc Pham, Siti Maryam Binte Ahmad Subaidi, S. Salleh, Shuo Sun, T. K. Vangani, Qiongqiong Wang, Won Cheng Yi Lewis, J. H. Wong, Jinyang Wu, Huayun Zhang, Longyin Zhang, Xunlong Zou
/ Abstract
We present MERaLiON-SER, a robust speech emotion recognition model designed for English and Southeast Asian languages. The model is trained using a hybrid objective combining weighted categorical cross-entropy and Concordance Correlation Coefficient (CCC) losses for joint discrete and dimensional emotion modelling. This dual approach enables the model to capture both the distinct categories of emotion (like happy or angry) and the fine-grained, such as arousal (intensity), valence (positivity/negativity), and dominance (sense of control), leading to a more comprehensive and robust representation of human affect. Extensive evaluations across multilingual Singaporean languages (English, Chinese, Malay, and Tamil ) and other public benchmarks show that MERaLiON-SER consistently surpasses both open-source speech encoders and large Audio-LLMs. These results underscore the importance of specialised speech-only models for accurate paralinguistic understanding and cross-lingual generalisation. Furthermore, the proposed framework provides a foundation for integrating emotion-aware perception into future agentic audio systems, enabling more empathetic and contextually adaptive multimodal reasoning.
Journal: ArXiv