Generative AI Model Learns Emotional Body Motion Expressions

Researchers have developed a Transformer-based generative model to learn emotional body motion expressions directly from motion-capture data. Trained on performances by Japanese actors, the model generates expressive motions conditioned on discrete emotion labels. Evaluations showed that machine observers could recognize emotions in the generated motions with 22.80% accuracy, while human raters achieved 24.91% accuracy. The study also demonstrated the model's utility in augmenting emotion recognition, extracting emotion-specific patterns, and synthesizing transitions between emotion intensities, highlighting its potential for affective computing. AI

IMPACT This research could enhance affective computing applications and improve the realism of virtual agents and social robots by enabling more nuanced emotional expression.

RANK_REASON The cluster contains an academic paper detailing a new research methodology and findings in AI. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• Emotional body motion expressions
• Generative model
• Japanese actors
• LSTM
• Transformer

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →