Metraipolis › archive › dc_45ca73a7bee4
AtomEgo: Exploring Ego-Robot Integration for Embodied Foundation Model Pretraining — arXiv:2609.21461
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.21461
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-10T01:10:32.037Z
- SHA-256 of the archived text
c16dbbe21c4a8df699af286d44c55db9aa23cff5f9b78dd8fdfd7a35c7965e4a- Stable URL of this page
https://metraipolis.com/documento/dc_45ca73a7bee4
How to cite
AtomEgo: Exploring Ego-Robot Integration for Embodied Foundation Model Pretraining — arXiv:2609.21461 — arXiv (en), https://arxiv.org/abs/2609.21461 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_45ca73a7bee4 (Metraipolis archive dc_45ca73a7bee4, sha256 c16dbbe21c4a8df699af286d44c55db9aa23cff5f9b78dd8fdfd7a35c7965e4a)
BibTeX
@misc{dc45ca73a7bee4,
title = {AtomEgo: Exploring Ego-Robot Integration for Embodied Foundation Model Pretraining — arXiv:2609.21461},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.21461},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_45ca73a7bee4; sha256 c16dbbe21c4a8df699af286d44c55db9aa23cff5f9b78dd8fdfd7a35c7965e4a},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.21461. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Embodied foundation models are constrained by the limited scale and diversity of robot demonstrations, motivating the use of large-scale egocentric human interaction data. However, how to effectively incorporate such data into embodied-model pre-training remains unclear because of substantial embodiment and action-space gaps between humans and robots. We present AtomEgo, a systematic study of ego--robot co-training supported by a curated corpus of approximately 2,659 hours and a scalable data processing pipeline. Across vision--language--action and world--action model architectures, we investigate three representative paradigms: joint co-training with domain-specific action heads, progressive ego-to-robot transfer through embodiment alignment, and joint video--action modeling. We evaluate these paradigms through multi-task real-robot experiments and language-conditioned cross-embodiment representation analysis. Our results reveal a simple principle: Data Scale * Alignment Quality --> Capability Gain; egocentric data can improve generalization, but their value depends on how effectively they are aligned and utilized. This principle can provide practical guidance for scalable ego--robot pre-training. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-21 in cs.AI). Categorie arXiv: Robotics (cs.RO) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2609.21461 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-10T00:23:42.538Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-21?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).