Research map / Label-efficient and robust learning / Recognition and long-tail
Generative End-to-End Autonomous Driving
Wenzhao Zheng, Ruiqi Song, Xianda Guo, Chenming Zhang, Long Chen
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 ego-to-agent interaction to VAD-tiny (ablation) | VAD-Tiny · nuScenes val planning | Avg. L2 (m): 1.30 -> 1.10 | Improved |
| add ego-to-agent interaction (self-attention over ego and agent tokens) (ablation) | GenAD · nuScenes val planning | Avg. Collision Rate (%): 1.22 -> 0.43 | Improved |
| add trajectory prior modeling (VAE latent space) (ablation) | GenAD · nuScenes val planning | Avg. L2 (m): 1.11 -> 1.02 | Improved |
| add latent future trajectory generation (GRU) (ablation) | GenAD · nuScenes val planning | Avg. Collision Rate (%): 0.70 -> 0.56 | Improved |
| combine trajectory prior modeling with latent GRU generation (ablation) | GenAD · nuScenes val planning | Avg. L2 (m): 1.02 -> 0.91 | Improved |
| trajectory prior modeling alone raises early-horizon collision rate (ablation) | GenAD · nuScenes val planning | Collision Rate 1s (%): 0.06 -> 0.23 | Hurt |
| use generative GenAD vs VAD-Tiny on planning | GenAD · nuScenes val planning | Avg. L2 (m): 1.30 -> 0.91 | Improved |
| GenAD vs UniAD on collision rate (UniAD uses extra tracking and occupancy supervision) | GenAD · nuScenes val planning | Avg. Collision Rate (%): 0.31 (UniAD) vs 0.43 (GenAD) | Hurt |
| GenAD vs VAD on perception and prediction | GenAD · nuScenes val detection | mAP: 0.27 -> 0.29 | Improved |
Closest papers by meaning
- Generating Multimodal Driving Scenes via Next-Scene Prediction (CVPR 2025 · similarity 0.90)
- Generative Scenario Rollouts for End-to-End Autonomous Driving (NeurIPS 2026 · similarity 0.90)
- Future-Aware End-to-End Driving: Bidirectional Modeling of Trajectory Planning and Scene Evolution (NeurIPS 2025 · similarity 0.89)
- Generalized Predictive Model for Autonomous Driving (CVPR 2024 · similarity 0.89)
- Scenario Dreamer: Vectorized Latent Diffusion for Generating Driving Simulation Environments (CVPR 2025 · similarity 0.89)
- Enhancing End-to-End Autonomous Driving with Latent World Model (ICLR 2025 · 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.
