A woman wearing a loose-fitting white blouse with ruffled sleeves, paired with high-waisted, wide-leg navy blue pants and black ankle-strap stiletto heels.

In this supplementary material, we first give additional details about the method in Sec. 1. Queries used for zero-shot semantic segmentation. We do this for all the annotated classes in the dataset (second column). One can see that, for example, class name'manmade' lacks descriptive specificity. In the text description of this class, we can find "... buildings, walls, guard rails, fences, poles, street signs, traffic lights ..." and more. Table 1: Queries used for zero-shot semantic segmentation.

The L-1 regression problem can be interpreted as maximum likelihood with a Laplacian error model. By taking negative log on Eq. 1, we have, ln P (t

Though achieving promising results, these methods may still suffer from several limitations.

Street-view imagery (SVI) offers a fine-grained lens on traffic risk, yet two fundamental challenges persist: (1) how to construct street-level indicators that capture accident-related features, and (2) how to quantify their causal impacts across different accident types. To address these challenges, we propose Semantic4Safety, a framework that applies zero-shot semantic segmentation to SVIs to derive 11 interpretable streetscape indicators, and integrates road type as contextual information to analyze approximately 30,000 accident records in Austin. Specifically, we train an eXtreme Gradient Boosting (XGBoost) multi-class classifier and use Shapley Additive Explanations (SHAP) to interpret both global and local feature contributions, and then apply Generalized Propensity Score (GPS) weighting and Average Treatment Effect (ATE) estimation to control confounding and quantify causal effects. Results uncover heterogeneous, accident-type-specific causal patterns: features capturing scene complexity, exposure, and roadway geometry dominate predictive power; larger drivable area and emergency space reduce risk, whereas excessive visual openness can increase it. By bridging predictive modeling with causal inference, Semantic4Safety supports targeted interventions and high-risk corridor diagnosis, offering a scalable, data-informed tool for urban road safety planning.

A woman wearing a loose-fitting white blouse with ruffled sleeves, paired with high-waisted, wide-leg navy blue pants and black ankle-strap stiletto heels.

In this supplementary material, we first give additional details about the method in Sec. 1. Queries used for zero-shot semantic segmentation. We do this for all the annotated classes in the dataset (second column). One can see that, for example, class name'manmade' lacks descriptive specificity. In the text description of this class, we can find "... buildings, walls, guard rails, fences, poles, street signs, traffic lights ..." and more. Table 1: Queries used for zero-shot semantic segmentation.

The L-1 regression problem can be interpreted as maximum likelihood with a Laplacian error model. By taking negative log on Eq. 1, we have, ln P (t

Though achieving promising results, these methods may still suffer from several limitations.

Additional Feedback: Here some detailed comments and questions: Fig 1a: is that label correct for 1a? Seems correct for 1b and 1c. L52: At this point nothing seems segmentation specific. Does this method also work for zero-shot classification? Maybe hint at what is pixel-level specific.