cs.RONov 10, 2026

Ego Guidance: Enhancing Robot Policy Learning with Action-Free Human Egocentric Videos

Authors: Lin Long, Tianren Ma, Zhong Chu, Di Wu, Maosong Cao, Qixiang Ye, Junbo Zhao, Shanghang Zhang

Organizations: Zhejiang University · Ant Group · University of Chinese Academy of Sciences · Nanyang Technology University · 

Abstract

In this paper, we investigate how human egocentric videos can benefit robot policy learning without action supervision. Building on a pretrained VLM, we propose Shared Prediction on Action and Representation (SPAR), an MoT architecture in which future visual state prediction can provide independent and complementary learning signals alongside action prediction. We find that representations learned from egocentric videos improve downstream performance on out-of-distribution robot tasks, with further gains as video data scales. We also study the effects of objective choices and data curation, providing insights into how to leverage action-free videos for scalable robot learning.

Figures & tables

Appendix figures & tables2 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. EgoForesight: Effectively Scaling Policy Learning with Noisy Egocentric Actions

    Oct 22, 2026Chubin Zhang, Changyuan Wang, Yirui Wang +7Robot Imitation LearningRobot Learning from Video

  2. EXE: Egocentric Cross-Embodiment Pretraining via Factorized Latent Actions for VLA Models

    Oct 24, 2026Qingyuan Cai, Bo Zhang, Yongzhen Huang +1Cross-Embodiment Robot LearningRobot Learning from Video

  3. Video2STL: Grounding VLM-Generated Temporal Specifications for Robot Learning

    Sep 29, 2026Merve Atasever, Keyan Azbijari, Cagan Bakirci +6VLMs for RoboticsSignal Temporal Logic