AI for Social Good: Advancing Autism Detection through Multi-Modal Emotion Estimation

Integrating Machine Learning into healthcare offers tremendous opportunities for early intervention and accessible care. This presentation highlights ongoing research at Universiti Malaya focusing on applying computer vision and deep learning for continuous emotion detection. We will discuss our recent work on classifying and monitoring Autism Spectrum Disorder (ASD) in children using predicted categorical and dimensional emotions from video features, alongside integrating vocal biomarkers. The session will emphasize how robust AI models can be practically deployed to support vulnerable demographics.