AD^2-Bench: A Hierarchical CoT Benchmark for MLLM in Autonomous Driving under Adverse Conditions

TL;DR

AD^2-Bench is a new hierarchical Chain-of-Thought benchmark designed to evaluate multi-modal large models' reasoning in autonomous driving under adverse weather and complex scenes, highlighting current models' limitations.

Contribution

It introduces the first comprehensive CoT benchmark for autonomous driving in challenging conditions, with extensive annotations and an evaluation framework for fine-grained reasoning analysis.

Findings

State-of-the-art MLLMs achieve below 60% accuracy on AD^2-Bench.

The benchmark reveals significant challenges in current models' reasoning capabilities.

AD^2-Bench facilitates targeted improvements in autonomous driving AI systems.

Abstract

Chain-of-Thought (CoT) reasoning has emerged as a powerful approach to enhance the structured, multi-step decision-making capabilities of Multi-Modal Large Models (MLLMs), is particularly crucial for autonomous driving with adverse weather conditions and complex traffic environments. However, existing benchmarks have largely overlooked the need for rigorous evaluation of CoT processes in these specific and challenging scenarios. To address this critical gap, we introduce AD^2-Bench, the first Chain-of-Thought benchmark specifically designed for autonomous driving with adverse weather and complex scenes. AD^2-Bench is meticulously constructed to fulfill three key criteria: comprehensive data coverage across diverse adverse environments, fine-grained annotations that support multi-step reasoning, and a dedicated evaluation framework tailored for assessing CoT performance. The core contribution of AD^2-Bench is its extensive collection of over 5.4k high-quality, manually annotated CoT instances. Each intermediate reasoning step in these annotations is treated as an atomic unit with explicit ground truth, enabling unprecedented fine-grained analysis of MLLMs' inferential processes under text-level, point-level, and region-level visual prompts. Our comprehensive evaluation of state-of-the-art MLLMs on AD^2-Bench reveals accuracy below 60%, highlighting the benchmark's difficulty and the need to advance robust, interpretable end-to-end autonomous driving systems. AD^2-Bench thus provides a standardized evaluation platform, driving research forward by improving MLLMs' reasoning in autonomous driving, making it an invaluable resource.