EUVIMEDDécouvrir EUVIMED
Open access · CC BY 4.0

GazeCapsNet: A Lightweight Gaze Estimation Framework

Shakhnoza Muksimova iD, Yakhyokhuja Valikhujaev iD, Sabina Umirzakova, Jushkin Baltayev, Young Im Cho iD

Sensors · 2025

DOI: 10.3390/s25041224 ↗

At a glance

Journal article in Sensors, 2025. Written by Shakhnoza Muksimova and 4 co-authors. Institution: Department of Computer Engineering, Gachon University, Sujeong-gu, Seongnam-si 461-701, Republic of Korea. Major MEDLINE topics: Fixation, Ocular, Eye-Tracking Technology. Cited 27 times so far (Crossref); 42 references. The full text is freely available (open access, CC BY 4.0). Funded by: Korean Agency for Technology and Standards under the Ministry of Trade, Industry, and Energy.

RésuméSubjectsRéférences bibliographiquesCiteRelated studies

Abstract (original language)

Gaze estimation is increasingly pivotal in applications spanning virtual reality, augmented reality, and driver monitoring systems, necessitating efficient yet accurate models for mobile deployment. Current methodologies often fall short, particularly in mobile settings, due to their extensive computational requirements or reliance on intricate pre-processing. Addressing these limitations, we present Mobile-GazeCapsNet, an innovative gaze estimation framework that harnesses the strengths of capsule networks and integrates them with lightweight architectures such as MobileNet v2, MobileOne, and ResNet-18. This framework not only eliminates the need for facial landmark detection but also significantly enhances real-time operability on mobile devices. Through the innovative use of Self-Attention Routing, GazeCapsNet dynamically allocates computational resources, thereby improving both accuracy and efficiency. Our results demonstrate that GazeCapsNet achieves competitive performance by optimizing capsule networks for gaze estimation through Self-Attention Routing (SAR), which replaces iterative routing with a lightweight attention-based mechanism, improving computational efficiency. Our results show that GazeCapsNet achieves state-of-the-art (SOTA) performance on several benchmark datasets, including ETH-XGaze and Gaze360, achieving a mean angular error (MAE) reduction of up to 15% compared to existing models. Furthermore, the model maintains a real-time processing capability of 20 milliseconds per frame while requiring only 11.7 million parameters, making it exceptionally suitable for real-time applications in resource-constrained environments. These findings not only underscore the efficacy and practicality of GazeCapsNet but also establish a new standard for mobile gaze estimation technologies.

Subjects

MeSH terms (MEDLINE)

Fixation, OcularEye-Tracking TechnologyHumansAlgorithmsAttentionVirtual RealityEye Movements

Author keywords

capsule networkseye appearancegaze estimationlightweight architecturesself-attention routing mechanism

Authors and institutions

  • Department of Computer Engineering, Gachon University, Sujeong-gu, Seongnam-si 461-701, Republic of Korea.
  • Aria Studios Co., Ltd., Seoul 06247, Republic of Korea.
  • Department of Information Systems and Technologies of the Tashkent State University of Economic, Tashkent 100066, Uzbekistan.

ORCID profiles: Shakhnoza Muksimova, Yakhyokhuja Valikhujaev, Young Im Cho

Funding

  • Korean Agency for Technology and Standards under the Ministry of Trade, Industry, and Energy

Cite

APA

Muksimova, S., Valikhujaev, Y., Umirzakova, S., Baltayev, J., & Cho, Y. I. (2025). GazeCapsNet: A Lightweight Gaze Estimation Framework. Sensors. https://doi.org/10.3390/s25041224

Vancouver

Muksimova S, Valikhujaev Y, Umirzakova S, Baltayev J, Cho YI. GazeCapsNet: A Lightweight Gaze Estimation Framework. Sensors. 2025. doi:10.3390/s25041224

BibTeX
@article{Muksimova2025,
  author = {Muksimova, Shakhnoza and Valikhujaev, Yakhyokhuja and Umirzakova, Sabina and Baltayev, Jushkin and Cho, Young Im},
  title = {GazeCapsNet: A Lightweight Gaze Estimation Framework},
  journal = {Sensors},
  year = {2025},
  doi = {10.3390/s25041224}
}
Start for free →