用场景元数据和任务提示提升自动驾驶问答的鲁棒性
Robust Driving QA through Metadata-Grounded Context and Task-Specific Prompts
- 分两阶段：先用多摄像头+历史帧生成推理链，再注入场景元数据和任务指令
- 结合自一致性集成后准确率达66.85%，比基线高4个百分点，严重视觉干扰下仍保持96%准确率
- 适合研究视觉语言模型在自动驾驶复杂决策中的应用，尤其关注提示工程与上下文增强
我们提出一个两阶段视觉语言问答系统，用于自动驾驶中的高层感知、预测与规划问题。第一阶段使用Qwen2.5-VL-32B大模型，输入六路摄像头图像、短时历史帧及带少样本示例的思维链提示，通过自一致性集成（多次采样推理链）提升答案可靠性。第二阶段在提示中加入nuScenes场景元数据（如物体标注、车辆状态）和任务特定指令（感知、预测、规划分别用不同提示）。在驾驶问答基准测试中，使用5帧历史和10次提示时，第一阶段准确率达65.1%（零样本为62.61%），自一致性提升至66.85%；第二阶段整体准确率达67.37%。尤其在严重视觉退化条件下仍保持96%准确率。结果表明，精心设计的提示与上下文锚定可显著增强预训练视觉语言模型在驾驶问答中的表现。
原文摘要 · Abstract (English)
We present a two-phase vision-language QA system for autonomous driving that answers high-level perception, prediction, and planning questions. In Phase-1, a large multimodal LLM (Qwen2.5-VL-32B) is conditioned on six-camera inputs, a short temporal window of history, and a chain-of-thought prompt with few-shot exemplars. A self-consistency ensemble (multiple sampled reasoning chains) further improves answer reliability. In Phase-2, we augment the prompt with nuScenes scene metadata (object annotations, ego-vehicle state, etc.) and category-specific question instructions (separate prompts for perception, prediction, planning tasks). In experiments on a driving QA benchmark, our approach significantly outperforms the baseline Qwen2.5 models. For example, using 5 history frames and 10-shot prompting in Phase-1 yields 65.1% overall accuracy (vs.62.61% with zero-shot); applying self-consistency raises this to 66.85%. Phase-2 achieves 67.37% overall. Notably, the system maintains 96% accuracy under severe visual corruption. These results demonstrate that carefully engineered prompts and contextual grounding can greatly enhance high-level driving QA with pretrained vision-language models.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。