Preprint
Oct 2026
DEFT-VLA: Stage-Localized Force Specialization for Contact-Rich Robotic Manipulation
DFT-VLA (Dual-Expert Force-specialized Training for Vision-Language-Action Manipulation) is proposed, which retains the task-adapted base expert for vision-dominated non-contact behavior and introduces a force-aware contact expert for critical contact stages, localizing force-conditioned adaptation to where it is neede...
Zan Huang, Yi-Bo Qiu, Kang-Li Ding et al.
· 0 citations