RESEARCH PAPERYear 2025

EgoBridge: Domain Adaptation for Generalizable Imitation from Egocentric Human Data

Ryan Punamiya; Dhruv Patel; Patcharapong Aphiwetsa; Pranav Kuppili; Lawrence Zhu; Simar Kareer; Judy Hoffman; Danfei Xu

Classification

Major category: Related resources
Quadrant: Not applicable
Architecture: Not applicable
Prediction paradigm: Not applicable
Subcategories: Action policy foundations Training optimization & distillation
Source review status: Verified from primary sources

Category review. EgoBridge aligns labeled human and robot observation/action distributions during imitation learning using DTW-guided optimal transport. Its deployed policy directly emits action chunks without language conditioning or future-world prediction. It belongs in Related resources as robot imitation/domain-adaptation methodology. Reading evidence

AT A GLANCE

Contribution

A contribution summary has not been added yet.

Abstract

An abstract has not been added yet.

Affiliations

Not listed in the collection.

BibTeX

@inproceedings{Punamiya2025EgoBridgeDomain,
 author = {Punamiya, Ryan and Patel, Dhruv and Aphiwetsa, Patcharapong and Kuppili, Pranav and Zhu, Lawrence and Kareer, Simar and Hoffman, Judy and Xu, Danfei},
 booktitle = {NeurIPS},
 pages = {47092--47122},
 title = {EgoBridge: Domain Adaptation for Generalizable Imitation from Egocentric Human Data},
 volume = {38},
 year = {2025}
}

Classification

Contribution

Abstract

Affiliations

BibTeX

More in this category

Valerant: An Automatic Navigable Game Map Generator via Action-Conditioned World Model Exploration

Toward Unified Robot Learning: Bridging Representation, Vision-Language-Action, and World Models

From World Models to World Action Models: A Concise Tutorial for Robotics