Research map / Video / Video generation
Taming Video Models for 3D and 4D Generation via Zero-Shot Camera Control
Chenxi Song, Yanming Yang, Tong Zhao, Ruibo Li, Chi Zhang
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 DSG (dual-path self-corrective guidance) (ablation) | WorldForge · 3D static scenes | FID: 109.43 -> 96.08 | Improved |
| add DSG (ablation) | WorldForge · 4D dynamic scenes | FVD: 95.69 -> 93.17 | Improved |
| add FLF (flow-gated latent fusion) (ablation) | WorldForge · 3D static scenes | FID: 112.69 -> 96.08 | Improved |
| add FLF (ablation) | WorldForge · 4D dynamic scenes | FVD: 99.79 -> 93.17 | Improved |
| add both DSG and FLF (ablation) | WorldForge · 4D dynamic scenes | FVD: 103.17 -> 93.17 | Improved |
| use orthogonal-component DSG instead of standard CFG formulation (ablation) | WorldForge · 3D static scenes | FID: 120.91 -> 96.08 | Improved |
| use orthogonal-component DSG instead of standard CFG formulation (ablation) | WorldForge · 4D dynamic scenes | FVD: 109.1 -> 93.17 | Improved |
| add IRR (intra-step recursive refinement) (ablation) | WorldForge · qualitative ablation | trajectory following | Improved |
| add FLF and DSG (ablation) | WorldForge · 3D static scenes | CLIPsim: 0.943 -> 0.948 | No clear effect |
| use WorldForge training-free guidance instead of training-based TrajectoryCrafter | Wan2.1 · 3D static scenes | FID: 111.49 -> 96.08 | Improved |
| use WorldForge instead of TrajectoryCrafter | Wan2.1 · 4D dynamic scenes | ATE: 0.431 -> 0.527 | Hurt |
| port the guidance to SVD and LongCat-Video | WorldForge · qualitative | visual quality | Improved |
Closest papers by meaning
- Training-free Camera Control for Video Generation (ICLR 2025 · similarity 0.89)
- Taming Video Diffusion Prior with Scene-Grounding Guidance for 3D Gaussian Splatting from Sparse Inputs (CVPR 2025 · similarity 0.89)
- Wan-Move: Motion-controllable Video Generation via Latent Trajectory Guidance (NeurIPS 2025 · similarity 0.89)
- AC3D: Analyzing and Improving 3D Camera Control in Video Diffusion Transformers (CVPR 2025 · similarity 0.88)
- Motion-Zero: A Zero-Shot Trajectory Control Framework of Moving Object for Diffusion-Based Video Generation (AAAI 2025 · similarity 0.88)
- Controlling Space and Time with Diffusion Models (ICLR 2025 · similarity 0.88)
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.
