technique

Force-Informed Re-Sampling Training

techniqueactiveprovisionalforce-informed-re-sampling-training-80324bbb·1 events·first seen 6d ago

Aliases: Force-Informed Re-Sampling Training

Co-occurring entities

Neural External Torque Estimation FACTR 2

More like this (12)

quantization-aware training defer-to-resample sparse frame sampling Quantisation-Aware Training (QAT)finite-sample posterior sampling framework UniIntervene: Agentic Intervention for Efficient Real-World Reinforcement Learning Reference-Augmented Training One-Shot Imitation Learning self-training shielded reinforcement learning Range Penalization: Theoretical Insights with Applications in Federated Learning Entropy-Regularized Reinforcement Learning

Recent events (1)

5arXiv · cs.LG·6d ago·source ↗

FACTR 2: Neural External Torque Estimation enables force-aware robot policy learning without dedicated sensors

Researchers introduce NEXT (Neural External Torque Estimation), a data-driven method that estimates external joint torques on commodity robot arms without dedicated force sensors, training in one minute from ten minutes of free-motion data. Combined with Force-Informed Re-Sampling Training (FIRST), which up-samples contact segments during behavior cloning, the system outperforms prior force-aware policies by over 17% in task progress across five long-horizon manipulation tasks. The work lowers the hardware barrier for contact-rich robot manipulation by bringing force-feedback teleoperation and policy learning to off-the-shelf arms.

Agent and Tool Ecosystem Neural External Torque Estimation Force-Informed Re-Sampling Training FACTR 2