Learning to See Like Humans: Gaze-Aligned Cycling Safety Prediction
For urban planners and researchers, this work improves the interpretability and accuracy of perception-based safety models by incorporating human visual attention, though the performance gain over existing methods is incremental.
The paper proposes EG-PCS, a framework that integrates eye-tracking data into a vision transformer-based pairwise learning pipeline to predict perceived cycling safety from street-view images. The gaze-guided model achieves ranking performance comparable to state-of-the-art methods while producing attention maps that better align with human visual attention.
Cycling delivers significant public-health and environmental benefits, yet its uptake in cities is often limited by perceived safety. When street environments appear unsafe, individuals are less likely to cycle, making perception a key barrier to adoption. Recent work has shown that pairwise comparisons of street-view images provide a scalable way to learn subjective safety judgments. However, existing approaches do not explicitly model human visual attention, which plays a central role in how humans perceive safety. We propose an Eye-Tracking-Guided Perceived Cycling Safety framework (EG-PCS) that integrates gaze data into a pairwise learning pipeline based on vision transformers. By supervising the model's attention mechanism with eye-tracking signals, we encourage alignment between learned attention maps and human fixation patterns. Experiments show that gaze-guided models achieve similar ranking performance compared to state-of-the-art approaches while producing attention maps that more accurately reflect human visual attention behavior. Our results demonstrate that incorporating eye-tracking information enhances both predictive accuracy and interpretability in perception-based urban analytics.