AI Paper Atlas

Research map / Video / Video generation

Composing Driving Worlds through Disentangled Control for Adversarial Scenario Generation

Yifan Zhan, Zhengqing Chen, Qingjie Wang, Zhuo He, Muyao Niu, Xiaoyang Guo, Wei Yin, Weiqiang Ren, Qian Zhang, zheng yinqiang

ECCV 2026 · Video › Video generation · approach: World-model video generation · 3 citations (OpenAlex)

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 changedWhereMeasuredEffect
use CompoSIA instead of MagicDrive-V2 for 6-view scene followingCompoSIA · nuScenes 6v scene followingFVD: 152.80 -> 133.66Improved
CompoSIA identity injection vs LoRA-Edit for identity controlCompoSIA · Identity controlFVD: 161.32 -> 149.15Improved
CompoSIA action conditioning vs MagicDrive-V2CompoSIA · Action followingTransErr (x1000): 13.66 -> 7.37Improved
CompoSIA vs MagicDrive-V2 under action editingCompoSIA · Action editingTransErr (x1000): 22.86 -> 12.15Improved
add local residual modulation (AdaLN) branch to action conditioning (ablation)CompoSIA · Action followingRotErr: 2.84 -> 0.55Improved
add local residual modulation (AdaLN) branch to action conditioning (ablation)CompoSIA · Action followingTransErr (x1000): 15.80 -> 7.37Improved
add global PRoPE attention branch to action conditioning (ablation)CompoSIA · Action followingTransErr (x1000): 11.24 -> 7.37Improved
add global PRoPE attention branch to action conditioning (ablation)CompoSIA · Action followingRotErr: 0.62 -> 0.55Improved
stop identity injection earlier (T_id=0.6) vs later (T_id=0.2) (ablation)CompoSIA identity injection · Identity control (qualitative)identity similarity vs editing freedomMixed
remove structure conditioning at sampling (ablation)CompoSIA · Qualitative ablationsurrounding vehicle motion consistencyImproved
inject identity cues throughout denoising rather than first-frame reference only (ablation)CompoSIA · Tunnel illumination scenes (qualitative)identity driftImproved
generated video instead of ground truth as planner inputEpona planner · nuScenes planningAvg collision rate (%): 0.35 -> 0.59Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most 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

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy