---
title: "LISA: Language-guided Interference-aware Spatial-Frequency Attention for Driver Gaze Estimation"
authors:
  - "Jun Ma"
  - "Zhenye Yang"
  - "Ruichen Zhou"
  - "Pei Zhang"
  - "Huan Li"
  - "Jinpeng Chen"
doi: "10.24963/ijcai.2026/325"
url: "https://ijcai.org/proceedings/2026/325"
pdf: "https://ijcai.org/proceedings/2026/0325.pdf"
venue: "Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence"
year: 2026
pages: "2924-2932"
track: "Main Track"
keywords:
  - "Data Mining: Applications"
  - "Machine Learning: Multi-modal learning"
---

# LISA: Language-guided Interference-aware Spatial-Frequency Attention for Driver Gaze Estimation

Jun Ma, Zhenye Yang, Ruichen Zhou, Pei Zhang, Huan Li, Jinpeng Chen

*Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence*, pages 2924-2932. DOI: [10.24963/ijcai.2026/325](https://doi.org/10.24963/ijcai.2026/325)

## Abstract

Driver gaze estimation serves as a fundamental metric for evaluating driver attentiveness in modern monitoring systems. Beyond being vulnerable to sudden lighting changes and sensor noise, spatial-domain models struggle to disentangle authentic gaze cues from irrelevant visual attributes. In this paper, we propose LISA, a Language-guided Interference-aware Spatial-Frequency Attention framework that combines frequency-domain priors with vision-language knowledge. Observing that the amplitude spectrum remains relatively stable even under spatial perturbations, we design a dual-domain fusion mechanism. It integrates stable low-frequency semantics into high-frequency details, employing spatial attention to precisely target ocular regions. To reduce semantic ambiguity, we also introduce a training-time disentanglement strategy. Using a frozen CLIP encoder and orthogonal regularization, we explicitly separate gaze features from appearance interference. Experiments on two benchmarks show that LISA achieves state-of-the-art performance, with significantly improved robustness against occlusions and lighting variations. The code repository is available at https://github.com/Mason-bupt/LISA.