Deep learning of dyadic interaction visual cues for human-robot collaboration in assembly tasks

Serving as a preliminary effort, these bodies of work seek to address the critical deficiencies in current dyadic human robot interaction methodologies, which typically depend on inadequate or isolated visual cues, failing to capture the dynamic and subtle human intentions essential for effective human-robot collaboration, especially within the framework of Industry 5.0.With a focus on dyadic interaction, this research establishes a foundational framework for transferring insights from human-to-human and human-to-robot surrogate interactions to facilitate effective human-robot collaborations, with novel contributions such as a gaze estimation model for subtle cue recognition and a multi-view dataset supporting advanced multi-view and intention-aware HRI frameworks.</p>