---
title: "PAC-MAN：面向全身安全的感知感知CBF-RL人形机器人躲球框架"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2607.28623v1"
permalink: "https://feed.astralor.com/items/itm_ca679b7835c324fca5e61440"
published: "2026-07-30T17:59:35.000Z"
score: 76
channel: "ai"
tags: ["PAC-MAN","Unitree","人形机器人","强化学习","感知","控制障碍函数"]
---

# PAC-MAN：面向全身安全的感知感知CBF-RL人形机器人躲球框架

## AI 点评

> 这篇论文让我特别兴奋的点在于，它把安全层从“全知全能”的假设中解放出来，直面真实传感器噪声。对做机器人部署的工程师来说，95%真实环境躲球成功率是个很硬的指标，说明这套方法不是只在仿真里好看。如果你正在做安全强化学习或人形机器人行为控制，这篇值得细读。

## 速览

该论文提出PAC-MAN框架，将控制障碍函数（CBF）安全保证与部署现实的机载感知结合，实现人形机器人全身躲球。策略仅通过头戴相机的分割掩码深度感知球体，训练时CBF引导提供全身链路安全间隔，并利用对抗运动先验正则化闪避反应。在可控任意链路接触基准上，单次投掷和部署循环中均接近特权状态预言机性能。研究发现CBF结构依赖于感知可观测性，联合CBF在准确球状态下最佳，在固定相机观测下退化，而球跟踪云台可恢复性能。最终在Unitree G1上零样本部署轻量级Link-CBF策略，真实世界95%成功率，并能通过语义分割躲避不同球体。

## 关键点

- 提出PAC-MAN框架，结合CBF安全约束与真实感知输入，实现人形机器人全身躲球。
- 仅使用头戴相机分割掩码深度信息，性能接近特权状态预言机。
- CBF结构性能依赖感知可观测性，固定相机下Joint-CBF退化，球跟踪云台可恢复。
- 在Unitree G1上零样本部署，真实世界95%成功率，支持语义分割躲避多种球体。

## 全文

We present PAC-MAN, a perception-aware CBF-RL framework that couples control-barrier safety with deployment-realistic onboard sensing for whole-body humanoid dodgeball. The deployed policy sees the ball only as segmentation-masked depth from a head-mounted camera, while training-time CBF guidance represents clearance to every body link, and an adversarial motion prior regularizes the resulting evasive reflexes. We evaluate on a controlled any-link contact benchmark with seeded throws in two regimes: single throws and a deployment loop in which the robot walks back to its station and recovers between throws. On this benchmark, the policy comes within a few points of a privileged state oracle: a fixed onboard camera alone is adequate for evasion. We find that usable barrier structure depends on perceptual observability: Joint-CBF gives the best performance with accurate ball states, degrades under fixed-camera observations when used only as training guidance, and recovers with a ball-tracking gimbal or privileged runtime filter. We therefore deploy a lightweight Link-CBF policy zero-shot on the Unitree G1 in the real world, where it tolerates imperfect perception, succeeds on 95% of throws, and uses semantic segmentation to dodge different balls.

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_ca679b7835c324fca5e61440