跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2611.13769 · v1

From Depth to Safe Depth: Uncertainty-Aware Monocular Perception for Mobile Robot Navigation

首次提交 2026-11-10 17:07
版本更新 2026-11-10 17:07
雷达观测 2026-11-11 11:04
arXiv ↗PDF ↗

摘要

Monocular depth foundation models make it tempting to replace depth sensors on mobile robots with a single RGB camera. We show that, used off the shelf, this is unsafe: in a simulated indoor navigation benchmark, a standard Nav2 stack driven by Depth Anything V2 (metric) collides in 63\% of 120 episodes, because the model systematically over-estimates the distance to nearby obstacles. We propose a lightweight plug-in module that turns predicted depth into \emph{safe depth}. A 0.58\,M-parameter head on top of the \emph{frozen} depth model predicts, per pixel, a log-depth correction μμ and an uncertainty σσ, trained with a ββ-weighted Gaussian negative log-likelihood. The navigation stack consumes the one-sided bound d^safe=d eμ−kσ\hat d_{\text{safe}}=d\,e^{μ-kσ}, where kk is set by conformal calibration on a held-out world, and uncertainty also decides which pixels may be treated as floor. The correction halves the depth error (AbsRel 0.262→0.1340.262\rightarrow0.134) and reduces near-range over-estimation from 59.8\% to 12.0\%. In navigation, safe depth raises the success rate from 37\% to 70\% and lowers collisions from 63\% to 3\%, close to the 2\% of ground-truth depth. A constant safety margin calibrated to the same pixel coverage reaches only 30\% success because it freezes the robot; uncertainty-aware safety wins 52 of 56 discordant paired episodes against it. Gains hold on glass, texture-less walls and dark regions.

AI 论文解读

基于摘要 · 非全文审阅

此版本尚未生成 AI 解读。解读由后台任务生成并按论文版本保存,登录后共享已有结果。

版本差异

目前只收录了一个版本。收录下一版后,可以对照标题、作者、分类和摘要。

观测到的版本

相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。

v1 · 2026-11-10 17:07 · 最新观测版本

From Depth to Safe Depth: Uncertainty-Aware Monocular Perception for Mobile Robot Navigation

Monocular depth foundation models make it tempting to replace depth sensors on mobile robots with a single RGB camera. We show that, used off the shelf, this is unsafe: in a simulated indoor navigation benchmark, a standard Nav2 stack driven by Depth Anything V2 (metric) collides in 63\% of 120 episodes, because the model systematically over-estimates the distance to nearby obstacles. We propose a lightweight plug-in module that turns predicted depth into \emph{safe depth}. A 0.58\,M-parameter head on top of the \emph{frozen} depth model predicts, per pixel, a log-depth correction μμ and an uncertainty σσ, trained with a ββ-weighted Gaussian negative log-likelihood. The navigation stack consumes the one-sided bound d^safe=d eμ−kσ\hat d_{\text{safe}}=d\,e^{μ-kσ}, where kk is set by conformal calibration on a held-out world, and uncertainty also decides which pixels may be treated as floor. The correction halves the depth error (AbsRel 0.262→0.1340.262\rightarrow0.134) and reduces near-range over-estimation from 59.8\% to 12.0\%. In navigation, safe depth raises the success rate from 37\% to 70\% and lowers collisions from 63\% to 3\%, close to the 2\% of ground-truth depth. A constant safety margin calibrated to the same pixel coverage reaches only 30\% success because it freezes the robot; uncertainty-aware safety wins 52 of 56 discordant paired episodes against it. Gains hold on glass, texture-less walls and dark regions.

查看此版本解读 →查看此版本原文 ↗