A new research paper evaluates existing gaze estimation methods for human-robot interaction (HRI) within a shared workspace. The study introduces a new annotated dataset collected using the NICO robotic platform and assesses four state-of-the-art gaze estimation models. While angular errors are comparable to general benchmarks, the practical limitations become apparent when translated to distance in the workspace, with the best median error reaching 14.57 cm. The paper concludes by discussing these limitations and providing recommendations for integrating gaze estimation into HRI systems. AI
IMPACT This research highlights practical limitations in current gaze estimation technology for human-robot interaction, suggesting areas for improvement in HRI system development.
RANK_REASON The cluster is about an academic paper detailing research findings and a new dataset. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →