Micro-behaviors Recognitions through Nonverbal Signals

Micro-behaviors are subtle, often unconscious actions, expressions, or gestures that individuals exhibit in their daily lives. They are often shown in nonverbal communication.

Understanding these subtle cues and endowing future artificial intelligence agents this capability could enhance our interactions and relationships. However, current automated systems cannot recognize micro-behaviors.

To this end, I will address this gap using multimodal perception and deep learning to automatically recognize micro-behaviors through nonverbal signals such as facial expressions and body language.

I will focus on the development of context-aware and privacy-preserving machine learning methods that are robust towards missing data.