# EgoForesight: Scaling Policy Learning with Noisy Egocentric Action Data

- Source: arXiv cs.RO
- Published: 2026-10-26T04:00:00.000Z
- RoboHot: https://robohot.news/en/items/fkd559ckcppadkftnbaah6vvs
- Original: https://arxiv.org/abs/2610.31845

## Summary

EgoForesight is a vision-language-action framework that jointly leverages robot demonstration trajectories and noisy egocentric videos for policy learning, reducing the impact of unreliable action labels through reliability-conditioned corruption and reliability-weighted regression, and supplementing supervision from demonstrated outcomes via privileged-future representation distillation.