Metraipolis › archive › dc_12956ce71869

XGenAct: Geometry-Enhanced World Action Models through Cross-Task Generation — arXiv:2610.03516

Source
arXiv
Original
https://arxiv.org/abs/2610.03516
License (as archived)
CC0-1.0
Archived
2026-10-05T23:10:56.215Z
SHA-256 of the archived text
98989475260e9ef1bdbe9a545a2217d639ab303d61f367dfe71700ed5c6608c8
Stable URL of this page
https://metraipolis.com/documento/dc_12956ce71869

How to cite

XGenAct: Geometry-Enhanced World Action Models through Cross-Task Generation — arXiv:2610.03516 — arXiv, https://arxiv.org/abs/2610.03516 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_12956ce71869 (Metraipolis archive dc_12956ce71869, sha256 98989475260e9ef1bdbe9a545a2217d639ab303d61f367dfe71700ed5c6608c8)
BibTeX
@misc{dc12956ce71869,
  title = {XGenAct: Geometry-Enhanced World Action Models through Cross-Task Generation — arXiv:2610.03516},
  howpublished = {arXiv},
  url = {https://arxiv.org/abs/2610.03516},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_12956ce71869; sha256 98989475260e9ef1bdbe9a545a2217d639ab303d61f367dfe71700ed5c6608c8},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.03516. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

World action models (WAMs) have advanced robot control by predicting how observations and actions evolve over time. Despite this progress, RGB and action based future prediction does not explicitly address the spatial understanding needed for robot manipulation. Existing efforts often add a limited set of spatial prediction tasks through specialized heads or branches, leaving both the range of spatial supervision and the model architecture fragmented. We introduce XGenAct, a world action model that represents RGB observations, robot actions, metric depth, surface normals, and functional role segmentation as RGB videos through deterministic codecs. By sampling perception and action tasks during training, XGenAct uses one video diffusion transformer and one objective to learn temporal prediction across these spaces without modality specific learned heads. On held out RLBench tasks, structured perception training improves average closed loop success over RGB only training, and XGenAct achieves 52% success in the five task external comparison, versus 26% for the strongest evaluated baselines. It also predicts future depth and segmentation more accurately than the evaluated pipelines that generate RGB first and then apply a frozen perception expert.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Robotics (cs.RO) ; Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.03516 — 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 2026-10-05T23:01:38.784Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).