Do You Know Where Your Camera Is? View-Invariant Policy Learning with Camera Conditioning

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Jiang, Tianchong, Ji, Jingtian, Tan, Xiangshan, Fang, Jiading, Bhattad, Anand, Guizilini, Vitor, Walter, Matthew R.
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866914071774756864
author Jiang, Tianchong
Ji, Jingtian
Tan, Xiangshan
Fang, Jiading
Bhattad, Anand
Guizilini, Vitor
Walter, Matthew R.
author_facet Jiang, Tianchong
Ji, Jingtian
Tan, Xiangshan
Fang, Jiading
Bhattad, Anand
Guizilini, Vitor
Walter, Matthew R.
contents We study view-invariant imitation learning by explicitly conditioning policies on camera extrinsics. Using Plucker embeddings of per-pixel rays, we show that conditioning on extrinsics significantly improves generalization across viewpoints for standard behavior cloning policies, including ACT, Diffusion Policy, and SmolVLA. To evaluate policy robustness under realistic viewpoint shifts, we introduce six manipulation tasks in RoboSuite and ManiSkill that pair "fixed" and "randomized" scene variants, decoupling background cues from camera pose. Our analysis reveals that policies without extrinsics often infer camera pose using visual cues from static backgrounds in fixed scenes; this shortcut collapses when workspace geometry or camera placement shifts. Conditioning on extrinsics restores performance and yields robust RGB-only control without depth. We release the tasks, demonstrations, and code at https://ripl.github.io/know_your_camera/ .
format Preprint
id arxiv_https___arxiv_org_abs_2510_02268
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Do You Know Where Your Camera Is? View-Invariant Policy Learning with Camera Conditioning
Jiang, Tianchong
Ji, Jingtian
Tan, Xiangshan
Fang, Jiading
Bhattad, Anand
Guizilini, Vitor
Walter, Matthew R.
Robotics
Computer Vision and Pattern Recognition
We study view-invariant imitation learning by explicitly conditioning policies on camera extrinsics. Using Plucker embeddings of per-pixel rays, we show that conditioning on extrinsics significantly improves generalization across viewpoints for standard behavior cloning policies, including ACT, Diffusion Policy, and SmolVLA. To evaluate policy robustness under realistic viewpoint shifts, we introduce six manipulation tasks in RoboSuite and ManiSkill that pair "fixed" and "randomized" scene variants, decoupling background cues from camera pose. Our analysis reveals that policies without extrinsics often infer camera pose using visual cues from static backgrounds in fixed scenes; this shortcut collapses when workspace geometry or camera placement shifts. Conditioning on extrinsics restores performance and yields robust RGB-only control without depth. We release the tasks, demonstrations, and code at https://ripl.github.io/know_your_camera/ .
title Do You Know Where Your Camera Is? View-Invariant Policy Learning with Camera Conditioning
topic Robotics
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2510.02268