Metraipolis › archive › dc_a25fd989cdf2
HuC-VideoMAE: Human-Centric Video Masked Autoencoding from synthetic data — arXiv:2610.08433
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.08433
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T13:07:42.165Z
- SHA-256 of the archived text
60b5bb7c9a02cc2170bea19ddf13c3e210badd7d7a0ce16259762ce489eb7f30- Stable URL of this page
https://metraipolis.com/documento/dc_a25fd989cdf2
How to cite
HuC-VideoMAE: Human-Centric Video Masked Autoencoding from synthetic data — arXiv:2610.08433 — arXiv (en), https://arxiv.org/abs/2610.08433 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_a25fd989cdf2 (Metraipolis archive dc_a25fd989cdf2, sha256 60b5bb7c9a02cc2170bea19ddf13c3e210badd7d7a0ce16259762ce489eb7f30)
BibTeX
@misc{dca25fd989cdf2,
title = {HuC-VideoMAE: Human-Centric Video Masked Autoencoding from synthetic data — arXiv:2610.08433},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.08433},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_a25fd989cdf2; sha256 60b5bb7c9a02cc2170bea19ddf13c3e210badd7d7a0ce16259762ce489eb7f30},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.08433. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Modern action recognition models rely on video transformers pretrained on massive collections of web-crawled videos, such as Kinetics-700. However, the use of such data raises ethical concerns, as subjects' consent is typically not obtained. Recent high-quality synthetic video datasets generated from motion-capture data, such as BEDLAM2.0, offer a promising ethical alternative. In this work, we investigate self-supervised pretraining of video transformers on synthetic human-motion datasets. We first show that directly applying the standard VideoMAE masking strategy leads to substantially worse performance than pretraining on Kinetics. To address this limitation, we propose a human-centric masking scheme that leverages body keypoints and person bounding box regions. Our approach encourages the model to focus on the structure and dynamics of human motion during pretraining. Experiments on NTU RGB+D and Toyota-Smarthome demonstrate that our method significantly outperforms standard VideoMAE pretraining on synthetic data, closing 49% of the gap to Kinetics pretraining on NTU RGB+D cross-view-subject without using a single real frame during pretraining. To promote the use of ethical action recognition models, we will publicly release our pretrained models. Tema degli esploratori: statistica e scienza dei dati (elenco arXiv cs.CV, nuovi lavori del giorno). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV). Pagina del lavoro: https://arxiv.org/abs/2610.08433 — 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 2026-10-07T12:37:32.989Z dalla pagina https://arxiv.org/list/cs.CV/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).