See Through the Noise: Improving Domain Generalization in Gaze Estimation

2026-04-17

AI Summary

Label noise significantly impairs the generalization of gaze estimation models in cross-domain scenarios. This work presents the first systematic investigation of this issue and introduces SeeTN, a novel framework that constructs a semantic embedding space to exploit the topological consistency between features and continuous labels. By designing a semantic manifold–based affinity regularization mechanism, SeeTN effectively identifies and mitigates label noise while preserving performance on the source domain. The approach integrates prototype transformation, domain-invariant relational modeling, and feature-label affinity measurement. Extensive experiments on multiple cross-domain gaze estimation benchmarks demonstrate that SeeTN substantially outperforms existing methods, achieving a favorable balance between denoising efficacy and source-domain accuracy.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Semi-Supervised Learning
• Natural Language Processing: Sentence-level Semantics, Textual Inference, etc.

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Generalizable gaze estimation methods have garnered increasing attention due to their critical importance in real-world applications and have achieved significant progress. However, they often overlook the effect of label noise, arising from the inherent difficulty of acquiring precise gaze annotations, on model generalization performance. In this paper, we are the first to comprehensively investigate the negative effects of label noise on generalization in gaze estimation. Further, we propose a novel solution, called See-Through-Noise (SeeTN) framework, which improves generalization from a novel perspective of mitigating label noise. Specifically, we propose to construct a semantic embedding space via a prototype-based transformation to preserve a consistent topological structure between gaze features and continuous labels. We then measure feature-label affinity consistency to distinguish noisy from clean samples, and introduce a novel affinity regularization in the semantic manifold to transfer gaze-related information from clean to noisy samples. Our proposed SeeTN promotes semantic structure alignment and enforces domain-invariant gaze relationships, thereby enhancing robustness against label noise. Extensive experiments demonstrate that our SeeTN effectively mitigates the adverse impact of source-domain noise, leading to superior cross-domain generalization without compromising the source-domain accuracy, and highlight the importance of explicitly handling noise in generalized gaze estimation.

Problem

• gaze estimation
• domain generalization
• label noise
• cross-domain generalization
• semantic structure

Innovation

• label noise
• domain generalization
• gaze estimation
• semantic embedding
• affinity regularization