Jul 2026
More Structure, Not More Capacity: Object-Centric Representations for Visuomotor Imitation Learning
An automated kinematic failure taxonomy separates spatial-precision (Near-Miss) failures from object-tracking (No-Grasp) failures: spatial grounding reduces Near-Miss while leaving No- Grasp unchanged.
Yi Li, Alexandre Chapin, Liming Chen et al.
· arXiv.org · 0 citations