Li-He Ding

RoboSynChallenge: Mastering Real-World Dexterity via Generalizing Synthesized Manipulation Skills

Achieving generalizable robotic manipulation remains a central challenge in embodied intelligence. Despite rapid advances in model architectures and learning algorithms, progress is often limited by the scarcity and narrow diversity of real-world data. The RoboSynChallenge competition introduces a unified benchmark to...