# RoadSceneBench: A Lightweight Benchmark for Mid-Level Road Scene Understanding

Source: [RoadSceneBench: A Lightweight Benchmark for Mid-Level Road Scene Understanding](<https://www.sota2.com/research/paper/692d217bd6b0dd3342931cba>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-12-02T05:06:24.125Z

Authors: Xiyan Liu, Han Wang, Yuhu Wang, Junjie Cai, Zhe Cao, Jianzhong Yang, Zhen Lu

Published: 2025-11-27T13:57:31.000Z

## Abstract

Understanding mid-level road semantics, which capture the structural and contextual cues that link low-level perception to high-level planning, is essential for reliable autonomous driving and digital map construction. However, existing benchmarks primarily target perception tasks such as detection or segmentation, overlooking the reasoning capabilities required to infer road topology and dynamic scene structure. To address this gap, we present RoadSceneBench, a lightweight yet information-rich benchmark designed to evaluate and advance visual reasoning in complex road environments. Unlike large-scale perception datasets, RoadSceneBench emphasizes relational understanding and structural consistency, encouraging models to capture the underlying logic of real-world road scenes. Furthermore, to enhance reasoning reliability, we propose Hierarchical Relational Reward Propagation with Temporal Consistency (HRRP-T), a training framework for Vision-Language Models (VLMs) in which reward signals adaptively promote spatial coherence and semantic alignment throughout the reasoning process. This paradigm enables models to move beyond static recognition toward geometry-aware and temporally consistent reasoning. Extensive experiments demonstrate that our method achieves state-of-the-art performance across diverse road configurations. RoadSceneBench thus provides a compact yet powerful foundation for studying mid-level road semantics and fostering structure-aware autonomous perception. Our dataset is available at https://github.com/XiyanLiu/RoadSceneBench.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Road Scene on RoadSceneBench](<https://www.sota2.com/research/sota/road-scene-on-roadscenebench.md>): Road Scene; RoadSceneBench; Precision: 82.8.

- [Ego-lane Index on RoadSceneBench](<https://www.sota2.com/research/sota/ego-lane-index-on-roadscenebench.md>): Ego-lane Index; RoadSceneBench; Precision (%): 75.44.

- [Lane Change Feasibility on RoadSceneBench](<https://www.sota2.com/research/sota/lane-change-feasibility-on-roadscenebench.md>): Lane Change Feasibility; RoadSceneBench; Precision: 87.63.

- [Lane Count on RoadSceneBench](<https://www.sota2.com/research/sota/lane-count-on-roadscenebench.md>): Lane Count; RoadSceneBench; Precision (%): 66.02.

- [Road Topology on RoadSceneBench](<https://www.sota2.com/research/sota/road-topology-on-roadscenebench.md>): Road Topology; RoadSceneBench; Precision: 0.6688.

- [Traffic Condition on RoadSceneBench](<https://www.sota2.com/research/sota/traffic-condition-on-roadscenebench.md>): Traffic Condition; RoadSceneBench; Precision: 82.35.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2511.22466>)

[PDF](<https://arxiv.org/pdf/2511.22466.pdf>)