Qwen-Drive-1.0：迈向自动驾驶视觉语言基础模型
原标题：Qwen-Drive-1.0: An Initial Step towards a Vision-Language Foundation Model for Autonomous Driving
AI 摘要
Qwen团队发布Qwen-Drive-1.0，这是首个面向自动驾驶的视觉语言基础模型，基于Qwen3.5-4B架构，集成3D感知、视觉问答和运动规划。模型通过外部BEV感知头和规划专家模块，在保持通用视觉语言能力的同时实现驾驶场景理解。实验显示其在3D感知和规划性能上具有竞争力，并已开源模型权重和代码。
正文节选
Qwen-Drive-1.0: An Initial Step towards a Vision-Language Foundation Model for Autonomous Driving Abstract We present Qwen-Drive-1.0, an initial step towards a vision-language foundation model for autonomous driving. Qwen-Drive-1.0 retains the architecture of the pretrained vision-language model (VLM) and integrates 3D perception, visual question answering, and motion planning within a unified framework. An external bird’s-eye-view (BEV) perception head jointly performs 3D object detection, sema