EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Data-Centric Pruning and Region-Weighted Loss for Robust Gaze Vector Estimation

Alejandro García de la Santa, Adrián Carrizo-Pérez, Ane Zulaika, Francisco J. Iriarte, Luis Unzueta, Iñigo Perona, Jose Luis Jodra, Arantxa Villanueva

IEEE Signal Processing Letters · 2026 · 33 · 1796-1800

DOI: 10.1109/lsp.2026.3682985 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Signal Processing Letters, 2026. Verfasst von Alejandro García de la Santa und 7 weiteren Autor:innen. Die Arbeit nennt 32 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Basque Government, ERDF A way of making Europe.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

In-the-wild eye tracking datasets to train gaze vector estimation methods are noisy and unreliable, due both to limited image quality and to weak guarantees on label accuracy. At the same time, simply increasing dataset size does not guarantee better results, especially when diversity, balance and label reliability are not taken into account. In this work, we adopt a data-centric perspective and introduce an innovative data-pruning pipeline for robust gaze vector estimation. Our methodology leverages vision-language model embeddings and their zero-shot classification capabilities, together with an early-learning prediction-error analysis, to assign task-aware quality scores to individual samples. Moreover, we incorporate a region-weighted loss conditioned on gaze direction to better handle the uneven difficulty across the gaze space and to improve learning in under-represented regions. Finally, we validate our approach on a large relabeled subset of the Gaze-Capture dataset by training a hybrid gaze-estimation model, and show that a curated subset of 84 K samples yields an accuracy improvement over a model trained on the full 2.45M-sample dataset.

Förderung

  • Basque Government
  • ERDF A way of making Europe

Zitieren

APA

Santa, A. G. D. L., Carrizo-Pérez, A., Zulaika, A., Iriarte, F. J., Unzueta, L., Perona, I., Jodra, J. L., & Villanueva, A. (2026). Data-Centric Pruning and Region-Weighted Loss for Robust Gaze Vector Estimation. IEEE Signal Processing Letters, 33, 1796-1800. https://doi.org/10.1109/lsp.2026.3682985

Vancouver

Santa AGDL, Carrizo-Pérez A, Zulaika A, Iriarte FJ, Unzueta L, Perona I, et al. Data-Centric Pruning and Region-Weighted Loss for Robust Gaze Vector Estimation. IEEE Signal Processing Letters. 2026;33:1796-1800. doi:10.1109/lsp.2026.3682985

BibTeX
@article{Santa2026,
  author = {Santa, Alejandro García de la and Carrizo-Pérez, Adrián and Zulaika, Ane and Iriarte, Francisco J. and Unzueta, Luis and Perona, Iñigo and Jodra, Jose Luis and Villanueva, Arantxa},
  title = {Data-Centric Pruning and Region-Weighted Loss for Robust Gaze Vector Estimation},
  journal = {IEEE Signal Processing Letters},
  year = {2026},
  volume = {33},
  pages = {1796--1800},
  doi = {10.1109/lsp.2026.3682985}
}
Kostenlos starten →