Research map / Vision-language and multimodal / Robot action and navigation
UniDrive-WM: Unified Understanding, Planning and Generation World Model For Autonomous Driving
Zhexiao Xiong, Xin Ye, Burhaneddin Yaman, Sheng Cheng, Yiren Lu, Jingru Luo, Nathan Jacobs, Liu Ren
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add future image generation head (ablation) | UniDrive-WM (AR) · Bench2Drive open-loop | L2 at 3s (m): 1.130 -> 1.079 | Improved |
| add future image generation head (ablation) | UniDrive-WM (AR) · Bench2Drive open-loop | Box collision at 3s (%): 0.746 -> 0.668 | Improved |
| add detection supervision (ablation) | UniDrive-WM (AR) · Bench2Drive open-loop | L2 at 3s (m): 2.146 -> 1.079 | Improved |
| add image generation to the model (ablation) | UniDrive-WM (AR) · Bench2Drive detection | NDS: 0.721 -> 0.746 | Improved |
| condition generation on planned trajectory (planning then generation) (ablation) | UniDrive-WM (AR) · Bench2Drive open-loop | FID: 9.1 -> 7.2 | Improved |
| condition generation on planned trajectory (planning then generation) (ablation) | UniDrive-WM (AR) · Bench2Drive open-loop | Avg. L2 (m): 0.67 -> 0.64 | Improved |
| add image generation (ablation) | UniDrive-WM · Chat-B2D VQA | CIDEr: 65.7 -> 66.7 | Improved |
| use AR+Diffusion branch instead of AR branch (ablation) | UniDrive-WM · Bench2Drive open-loop | L2 at 3s (m): 1.079 -> 1.066 | Improved |
| use AR+Diffusion branch instead of AR branch (ablation) | UniDrive-WM · Bench2Drive future frame generation | FID: 7.2 -> 6.6 | Improved |
| unified planning and generation vs ORION | UniDrive-WM (AR) · nuScenes open-loop | Avg. collision (%): 0.31 vs FSDrive 0.19; L2 0.30 vs ORION 0.34 | Mixed |
| unified model vs ORION on closed-loop | UniDrive-WM (AR) · Bench2Drive base set (closed-loop) | Driving Score: 77.74 -> 79.22 | Improved |
Later papers that compared with this method
1 later accepted paper reports UniDrive-WM as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- MindDrive: A Vision-Language-Action Model for Autonomous Driving via Online Reinforcement Learning (ECCV 2026 · better on 2 of 2 shared results)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- ORION — ORION: A Holistic End-to-End Autonomous Driving Framework by Vision-Language Instructed Action Generation (ICCV 2025) · better on 8 of 8 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Driving into the Future: Multiview Visual Forecasting and Planning with World Model for Autonomous Driving (CVPR 2024 · similarity 0.91)
- FutureSightDrive: Thinking Visually with Spatio-Temporal CoT for Autonomous Driving (NeurIPS 2025 · similarity 0.90)
- Future-Aware End-to-End Driving: Bidirectional Modeling of Trajectory Planning and Scene Evolution (NeurIPS 2025 · similarity 0.90)
- Driving in the Occupancy World: Vision-Centric 4D Occupancy Forecasting and Planning via World Models for Autonomous Driving (AAAI 2025 · similarity 0.90)
- SGDrive: Scene-to-Goal Hierarchical World Cognition for Autonomous Driving (CVPR 2026 · similarity 0.89)
- ChainFlow-VLA: Causal Flow Planning with Vision-Language Models (NeurIPS 2026 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
