系統將 3D BEV 感知同軌跡規劃設為外接模組,提供 VQA、直接規劃及先推理後規劃三種運行方式。
項目公布嘅基準成績可作研究參考,但屬項目自行報告,並非自動駕駛安全認證或可直接部署嘅證明。
權重喺 Hugging Face 提供;GitHub 則有推理程式、示範資料、文件同評測工具。
What is Alibaba Qwen’s Qwen-Drive-1.0-4B, released on September 8, 2026, and how does its Apache 2.0 open-source design combine the unchangeQwen-Drive-1.0-4B combines a shared vision-language backbone with separate perception and planning modules.
AI 提示
Create a landscape editorial hero image for this Studio Global article: What is Alibaba Qwen’s Qwen-Drive-1.0-4B, released on September 8, 2026, and how does its Apache 2.0 open-source design combine the unchange. Article summary: Qwen-Drive-1.0-4B is Qwen’s Apache-2.0 open-weight, 4B-parameter vision-language driving model, developed with Huazhong University of Science and Technology. It keeps Qwen3.5-4B’s multimodal VLM architecture intact and a. Topic tags: general, general web. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks, charts with fake numbers, clic