Event
Paper on GTP-FA Two-Stage Grasp-Planning Framework Released
1 reports · 1 sources
Jiahao Xu and other authors propose GTP-FA, a two-stage "grasp first, then plan" framework for long-horizon robot manipulation: it first generates grasp candidates, then performs downstream motion planning conditioned on the selected grasp. The framework also learns a failure attribution model that outputs a failure-mode distribution for failed trajectories, and uses it to separately optimize the grasp scoring and planning modules. The authors report that simulation and real-robot experiments show GTP-FA improves overall task success rates over baselines including RL, IL, diffusion policies, and VLA. The paper has been accepted to NeurIPS 2026.
Report timeline
- GTP-FA: A Two-Stage Grasp-Then-Plan Manipulation Framework with Failure Attribution
The authors propose GTP-FA, a two-stage "grasp-then-plan" framework for robotic manipulation: it first generates grasp candidates, then performs downstream motion planning conditioned on the selected grasp. The framework learns a failure attribution model that outputs a distribution over failure modes for failed trajectories, and uses it to separately optimize the grasp scoring and planning modules. Simulation and real-robot experiments show that GTP-FA improves overall task success rates over baselines including RL, IL, diffusion policies, and VLA, and the paper has been accepted to NeurIPS 2026.
arXiv cs.RO ·