Announcement_28

[NeurIPS 2026 E&D] Our paper “HandEdit: A Unified Benchmark for Egocentric Human-to-Robot Dexterous Hand Image Editing” has been accepted to the NeurIPS 2026 E&D Track. HandEdit introduces a large-scale egocentric human-to-robot dexterous-hand image-editing dataset and benchmark, covering over 200M editing instances and 26 robotic embodiments with an embodiment-aware evaluation protocol.