EUVIMEDEUVIMED entdecken
Crossref

On Model and Data Scaling for Skeleton-based Self-Supervised Gait Recognition

Adrian Cosma, Andy Eduard Catruna, Emilian Radoi

Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (5) · 3434-3442

DOI: 10.1609/aaai.v40i5.37340 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Adrian Cosma und 2 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Gait recognition from video streams is a challenging problem in computer vision biometrics due to the subtle differences between gaits and numerous confounding factors. Recent advancements in self-supervised pretraining have led to the development of robust gait recognition models that are invariant to walking covariates. While neural scaling laws have transformed model development in other domains by linking performance to data, model size, and compute, their applicability to gait remains unexplored. In this work, we conduct the first empirical study scaling on skeleton-based self-supervised gait recognition to quantify the effect of data quantity, model size and compute on downstream gait recognition performance. We pretrain multiple variants of GaitPT -- a transformer-based architecture -- on a dataset of 2.7 million walking sequences collected in the wild. We evaluate zero-shot performance across four benchmark datasets to derive scaling laws for data, model size, and compute. Our findings demonstrate predictable power-law improvements in performance with increased scale and confirm that data and compute scaling significantly influence downstream accuracy. We further isolate architectural contributions by comparing GaitPT with GaitFormer under controlled compute budgets. These results provide practical insights into resource allocation and performance estimation for real-world gait recognition systems.

Zitieren

APA

Cosma, A., Catruna, A. E., & Radoi, E. (2026). On Model and Data Scaling for Skeleton-based Self-Supervised Gait Recognition. Proceedings of the AAAI Conference on Artificial Intelligence, 40(5), 3434-3442. https://doi.org/10.1609/aaai.v40i5.37340

Vancouver

Cosma A, Catruna AE, Radoi E. On Model and Data Scaling for Skeleton-based Self-Supervised Gait Recognition. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(5):3434-3442. doi:10.1609/aaai.v40i5.37340

BibTeX
@article{Cosma2026,
  author = {Cosma, Adrian and Catruna, Andy Eduard and Radoi, Emilian},
  title = {On Model and Data Scaling for Skeleton-based Self-Supervised Gait Recognition},
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2026},
  volume = {40},
  number = {5},
  pages = {3434--3442},
  doi = {10.1609/aaai.v40i5.37340}
}
Kostenlos starten →