Polarization-Based Eye Tracking with Personalized Siamese Architectures

TL;DR

This paper demonstrates that Siamese architectures can personalize polarization-based eye tracking, significantly reducing calibration effort and improving accuracy compared to traditional methods.

Contribution

It introduces a Siamese-based personalization method for polarization eye tracking, achieving high accuracy with fewer calibration samples and outperforming NIR-based approaches.

Findings

Achieves comparable performance to linear calibration with 10 times fewer samples.

Reduces gaze error by up to 12% using polarization inputs over NIR inputs.

Combining Siamese personalization with linear calibration improves accuracy by up to 13%.

Abstract

Head-mounted devices integrated with eye tracking promise a solution for natural human-computer interaction. However, they typically require per-user calibration for optimal performance due to inter-person variability. A differential personalization approach using Siamese architectures learns relative gaze displacements and reconstructs absolute gaze from a small set of calibration frames. In this paper, we benchmark Siamese personalization on polarization-enabled eye tracking. For benchmarking, we use a 338-subject dataset captured with a polarization-sensitive camera and 850 nm illumination. We achieve performance comparable to linear calibration with 10-fold fewer samples. Using polarization inputs for Siamese personalization reduces gaze error by up to 12% compared to near-infrared (NIR)-based inputs. Combining Siamese personalization with linear calibration yields further improvements of up to 13% over a linearly calibrated baseline. These results establish Siamese personalization as a practical approach enabling accurate eye tracking.