跳到正文
arXiv cs.AI· Shiwei Wu, Yuwen Pan, Zhi Li, Yaowen Li, Yan Zhou, Xuemiao Xu, Yu Liu·· 10 小时前AI 评分38

AeroRSI:面向空中视觉语言导航的递归自我改进框架

AeroRSI: Recursive Self-Improvement for Aerial Vision-Language Navigation

AI 导读

AeroRSI 是一个让空中导航策略从自身失败中迭代学习的递归自我改进框架,核心思路是让失败策略通过"继续飞行"来验证修复动作是否可靠。在 TravelUAV 上,三轮迭代将 2B 参数监督策略的成功率从 46.4% 提升至 67.5%,碰撞减少一半,较同初始化的此前 SOTA 高出 10.6 个百分点,且无需额外人工标注或推理成本。代码与模型已公开。

正文 · 原文

View PDF HTML (experimental)

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 this https URL.
Comments: 8-page main paper with references and supplementary material (15 pages); 5 figures, 13 tables. Code: this https URL Project page: this https URL
Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Robotics (cs.RO)
Cite as: arXiv:2611.24077 [cs.CV]
  (or arXiv:2611.24077v1 [cs.CV] for this version)
  https://doi.org/10.48550/arXiv.2611.24077

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Shiwei Wu [view email]
[v1] Wed, 18 Nov 2026 09:07:04 UTC (5,231 KB)

来源:arXiv cs.AI · arxiv.org