Raqamli Transformatsiya va Sun’iy Intellekt
HAS MULTIMODAL LEARNING SUCCEEDED ENOUGH TO CAPTURE CONTEXTUAL MEANING OF HUMAN-TO-HUMAN INTERACTION? A SURVEY
1 Tashkent University of Information Technologies named after Muhammad al-Khwarizmi.
2 Tashkent State University of Economics.
Abstract
Human communication is inherently multimodal, involving speech, facial expressions, gestures, body language, and even contextual cues. Variability and ambiguity make more complex to understand contextual meaning of human-to-human interaction (HHI) as gestures and expressions may have different meanings across cultures and personal habits and styles influence behavior interpretation. To tackle these problems, this article systematically analyses past and current state-of-the-art researches in multimodal learning techniques for contextual understanding of HHI using audio, text, and vision data.
References
No references listed for this article.