Research map / Video / Video generation
NoiseController: Towards Consistent Multi-view Video Generation via Noise Decomposition and Collaboration
Haotian Dong, Xin Wang, Di Lin, Yipeng Wu, Qin Chen, Ruonan Liu, Kairui Yang, Ping Li, Qing Guo
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 joint denoising only (ablation) | MagicDrive baseline · nuScenes | FVD: 177.3 -> 173.6 | Improved |
| add multi-level noise decomposition (ablation) | joint denoising · nuScenes | FVD: 173.6 -> 166.1 | Improved |
| add multi-frame noise collaboration (ablation) | joint denoising · nuScenes | FVD: 173.6 -> 137.2 | Improved |
| add multi-frame noise collaboration to decomposition (ablation) | NoiseController · nuScenes | FVD: 166.1 -> 122.9 | Improved |
| add multi-level noise decomposition to collaboration (ablation) | NoiseController · nuScenes | FID: 15.36 -> 14.65 | Improved |
| add learnable spatiotemporal matrix S (ablation) | NoiseController · nuScenes | FVD: 154.4 -> 122.9 | Improved |
| make S learnable instead of fixed (ablation) | NoiseController · nuScenes | FVD: 183.4 -> 122.9 | Improved |
| add learnable impact matrix I (ablation) | NoiseController · nuScenes | FVD: 128.1 -> 122.9 | Improved |
| make I learnable instead of fixed (ablation) | NoiseController · nuScenes | FVD: 135.8 -> 122.9 | Improved |
| add NoiseController to MagicDrive | MagicDrive · nuScenes | FVD: 177.3 -> 122.9 | Improved |
| add NoiseController to TrackDiffusion | TrackDiffusion · nuScenes | FVD: 605.0 -> 590.2 | Improved |
| add NoiseController-generated data | BEVFusion · nuScenes | NDS: 37.61 -> 43.54 | Improved |
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.
- Panacea — Panacea: Panoramic and Controllable Video Generation for Autonomous Driving (CVPR 2024) · better on 2 of 2 shared results
- DriveDreamer — DriveDreamer: Towards Real-world-driven World Models for Autonomous Driving (ECCV 2024) · 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
- FreeLong: Training-Free Long Video Generation with SpectralBlend Temporal Attention (NeurIPS 2024 · similarity 0.87)
- How I Warped Your Noise: a Temporally-Correlated Noise Prior for Diffusion Models (ICLR 2024 · similarity 0.86)
- WorldWeaver: Generating Long-Horizon Video Worlds via Rich Perception (NeurIPS 2025 · similarity 0.85)
- Go-with-the-Flow: Motion-Controllable Video Diffusion Models Using Real-Time Warped Noise (CVPR 2025 · similarity 0.85)
- Ouroboros-Diffusion: Exploring Consistent Content Generation in Tuning-free Long Video Diffusion (AAAI 2025 · similarity 0.85)
- Dynamic View Synthesis as an Inverse Problem (NeurIPS 2025 · similarity 0.84)
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.
