Research map / Label-efficient and robust learning / Recognition and long-tail
ResWorld: Temporal Residual World Model for End-to-End Autonomous Driving
Jinqing Zhang, Zehua Fu, zelinxu, wenying.dai, Qingjie Liu, Yunhong Wang
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 TR-World and FGTR to the SSR-style baseline (no ego status) (ablation) | ResWorld · nuScenes | avg L2 / collision rate: L2 0.71 -> 0.65; CR 0.31 -> 0.23 | Improved |
| Add TR-World and FGTR (with ego status) (ablation) | ResWorld · nuScenes | avg L2 / collision rate: L2 0.65 -> 0.59; CR 0.28 -> 0.17 | Improved |
| Add TR-World only (ablation) | baseline with ego status · nuScenes | avg collision rate (%): 0.28 -> 0.25 | Improved |
| Add FGTR only (ablation) | baseline with ego status · nuScenes | avg collision rate (%): 0.28 -> 0.22 | Improved |
| Use TR-World instead of normal world model (ablation) | ResWorld · nuScenes | avg collision rate (%): 0.21 -> 0.17 | Improved |
| Add real future-data supervision of future BEV features to TR-World (ablation) | TR-World · nuScenes | avg collision rate (%): 0.17 -> 0.21 | Hurt |
| Add future supervision to normal world model (ablation) | Normal world model · nuScenes | avg collision rate (%): 0.21 -> 0.23 | No clear effect |
| Use ResWorld with historical-frame temporal residuals instead of agent queries | ResWorld · NAVSIM navtest | PDMS: 88.3 -> 89.0 | Improved |
| Final trajectory after FGTR refinement vs prior trajectory (ablation) | ResWorld · nuScenes | avg L2 / collision rate: L2 0.61 -> 0.59; CR 0.18 -> 0.17 | Improved |
Later papers that compared with this method
1 later accepted paper reports ResWorld as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Trajectory Evaluation via Rollout-Free World Model for End-to-end Autonomous Driving (NeurIPS 2026 · better on 1 of 1 shared result)
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.
- ST-P3 — ST-P3: End-to-End Vision-Based Autonomous Driving via Spatial-Temporal Feature Learning (ECCV 2022) · better on 2 of 2 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
- Future-Aware End-to-End Driving: Bidirectional Modeling of Trajectory Planning and Scene Evolution (NeurIPS 2025 · similarity 0.90)
- DriveDreamer4D: World Models Are Effective Data Machines for 4D Driving Scene Representation (CVPR 2025 · similarity 0.89)
- Driving in the Occupancy World: Vision-Centric 4D Occupancy Forecasting and Planning via World Models for Autonomous Driving (AAAI 2025 · similarity 0.89)
- Driving into the Future: Multiview Visual Forecasting and Planning with World Model for Autonomous Driving (CVPR 2024 · similarity 0.88)
- DrivingSphere: Building a High-fidelity 4D World for Closed-loop Simulation (CVPR 2025 · similarity 0.88)
- Enhancing End-to-End Autonomous Driving with Latent World Model (ICLR 2025 · similarity 0.87)
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.
