Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Minimally invasive prediction of blood lactate during incremental exercise via heart rate, core body temperature, and sweat-derived indices.

Scientific reports·2026

Same author

Elucidation of Enhanced Lysine Sensing via Pt-Functionalized 2D WS<sub>2</sub> for Biosensing Applications.

ACS applied materials & interfaces·2026

Same author

PD-L1 mRNA expression correlates with tumor growth rate in giant cell tumor of bone: a volumetric MRI analysis.

BMC musculoskeletal disorders·2026

Same author

Diffusion-based skin disease data augmentation with fine-grained detail preservation and interpolation for data diversity.

PloS one·2025

Same author

Correction: Discovery of genes positively modulating treatment effect using potential outcome framework and Bayesian update.

BMC medical informatics and decision making·2025

Same author

Selective denoising autoencoder for classification of noisy gas mixtures using 2D transition metal dichalcogenides.

Talanta·2024

Search research articles

Related Experiment Video

Updated: Sep 20, 2025

A Methodology for Capturing Joint Visual Attention Using Mobile Eye-Trackers

A Methodology for Capturing Joint Visual Attention Using Mobile Eye-Trackers

Published on: January 18, 2020

Improved Feature-Based Gaze Estimation Using Self-Attention Module and Synthetic Eye Images.

Jaekwang Oh¹, Youngkeun Lee¹, Jisang Yoo¹

¹Department of Electronic Engineering, Kwangwoon University, Seoul 01897, Korea.

Sensors (Basel, Switzerland)

|June 10, 2022

Summary

This study introduces a novel gaze estimation method using eye region landmarks, improving accuracy in low-resolution and noisy images. The approach achieves state-of-the-art performance, outperforming existing methods in real-world settings.

Keywords:

eye landmark detection gaze estimation based on feature self-attention synthetic eye images

More Related Videos

Gaze in Action: Head-mounted Eye Tracking of Children's Dynamic Visual Attention During Naturalistic Behavior

Gaze in Action: Head-mounted Eye Tracking of Children's Dynamic Visual Attention During Naturalistic Behavior

Published on: November 14, 2018

Eye Tracking During A Complex Aviation Task For Insights Into Information Processing

Eye Tracking During A Complex Aviation Task For Insights Into Information Processing

Published on: April 4, 2025

Related Experiment Videos

Last Updated: Sep 20, 2025

A Methodology for Capturing Joint Visual Attention Using Mobile Eye-Trackers

A Methodology for Capturing Joint Visual Attention Using Mobile Eye-Trackers

Published on: January 18, 2020

Gaze in Action: Head-mounted Eye Tracking of Children's Dynamic Visual Attention During Naturalistic Behavior

Gaze in Action: Head-mounted Eye Tracking of Children's Dynamic Visual Attention During Naturalistic Behavior

Published on: November 14, 2018

Eye Tracking During A Complex Aviation Task For Insights Into Information Processing

Eye Tracking During A Complex Aviation Task For Insights Into Information Processing

Published on: April 4, 2025

Area of Science:

Computer Vision
Machine Learning
Human-Computer Interaction

Background:

Gaze estimation is crucial for understanding user intent and attention.
Current deep learning methods often struggle with low-resolution and noisy images in unconstrained environments.
Appearance-based methods can yield unsatisfactory gaze estimation results under challenging real-world conditions.

Purpose of the Study:

To propose a novel gaze estimation method robust to low-resolution and noisy images.
To improve the accuracy and reliability of gaze estimation in unconstrained settings.
To achieve competitive or superior performance compared to existing appearance-based methods.

Main Methods:

Gaze estimation via detecting eye region landmarks from a single eye image.
Utilizing a large number of landmarks, including iris and eye edges, for rich information extraction.
Employing the HRNet backbone for robust feature learning at various resolutions.
Integrating the CBAM self-attention module for refined feature maps and enhanced spatial information.
Inputting extracted landmarks into a lightweight neural network for gaze prediction.

Main Results:

Achieved a landmark localization error of 3.18%, a 4% improvement over existing methods.
Demonstrated state-of-the-art gaze estimation performance of 4.32 degrees on the MPIIGaze dataset.
Showcased a 6% performance improvement compared to current benchmarks in naturalistic environments.
The proposed method exhibits enhanced robustness to noisy inputs and varying lighting conditions.

Conclusions:

The landmark-based gaze estimation method is competitive with appearance-based approaches.
The method offers significant improvements in accuracy and robustness for real-world gaze tracking.
This approach provides a reliable solution for gaze estimation in challenging, unconstrained environments.