Open Source
AeroRSI: Recursive Self-Improvement for Aerial Vision-Language Navigation
arXiv cs.RO0 views
arXiv:2611.24077v1 Announce Type: cross Abstract: Vision-language-action (VLA) models enable unmanned aerial vehicles (UAVs) to follow natural-language instructions, yet still fail frequently in closed-loop flight. Correcting these failures requires repairs, i.e., actions that avoid them at the states the policy itself reaches, which human pilots cannot label at scale and rule-based planners cannot supply reliably. In this paper, we present AeroRSI, a recursive self-improvement framework in which an aerial navigation policy iteratively learns from its own failures. Our key insight is that a failing policy should judge its own repairs by flying on from them. Judged in this way, its sampled actions and even rule-based proposals that hurt it when imitated directly become reliable supervision. Moreover, an episode terminates at its first failure, concealing subsequent failures until the earlier ones are repaired. AeroRSI thus renews its supervision in every round, anchoring the failures of the current policy, verifying repairs by its continuation, and distilling them while keeping its behavior elsewhere. On TravelUAV, three rounds raise the success rate of a supervised 2B-parameter policy from 46.4% to 67.5% and reduce collisions by half, surpassing the previous state of the art by 10.6 points from the same initialization without extra human labels or inference cost. Code and models are available at https://github.com/starsholic/AeroRSI.
Source:https://arxiv.org/abs/2611.24077PreviousAttention Is Not Relevance: Challenging the Causality of Attention in Autonomous DrivingNextObject Pose: The Link between Weak Perspective, Paraperspective, and Full Perspective
Related posts
DrawScore: Evaluating Generated CAD Models against Engineering Drawings
0arXiv cs.SE
Predicting Process-Level Behavior: A Benchmarking Study
0arXiv cs.SE
Characterizing GitHub Security Actions Alerts For Web Application Tutorials
0arXiv cs.SE
RoboBuilder: Agentic Robot Co-Design with Real-world Video Feedback in the Loop
0arXiv cs.RO
LightSEAD: Lightweight SE(3)-Aware Keypoint Detector through Selective Knowledge Distillation
0arXiv cs.RO
Render the Robot, Generate the World: One Video World Model Across Embodiments
0arXiv cs.RO