معرفی
Muhammad Usman Malik is a Researcher at the L3i laboratory (Laboratoire des sciences et techniques de l'information, de la communication et de la connaissance) at the University of La Rochelle. His work focuses on human-machine interaction, particularly in developing multimodal interaction models for mixed societies. He presented his thesis work at the IC meeting on September 3, 2021, detailing advancements in addressed detection, turn change prediction, and visual focus of attention generation in multiparty interactions.
Research Interests: Malik’s research integrates machine learning techniques to model complex social interactions. His work emphasizes the integration of verbal and non-verbal modalities (e.g., speech, gestures, gaze) to create more natural human-agent interactions. Key applications include predicting interaction dynamics in multi-participant scenarios and generating realistic visual attention behaviors for artificial agents.
Academic Contributions: His 2021 thesis introduced novel machine learning frameworks for addressed detection (using multiclass classification), turn change prediction (binary classification), and VFOA behavior generation (hybrid machine learning and heuristic methods). These models outperformed baseline approaches, with VFOA results achieving near-human-like perception in naturalness.
Labs/Teams: Affiliated with the L3i lab, Malik contributes to interdisciplinary projects at the University of La Rochelle. His work aligns with initiatives in AI, robotics, and human-computer interaction.
