Research map / Video / Video generation
CameraAnything: Refilming Videos with Arbitrary Camera Control
Yixuan Li, Yanhong Zeng, Ka Leong Cheng, Jiayi Zhu, Hanlin Wang, Wen Wang, Yihao Meng, Hao Ouyang, Qiuyu Wang, Yue Yu, Zidong Wang, Yiyuan Zhang and 2 more
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 |
|---|---|---|---|
| use Plucker rays with AdaLN instead of linear camera vector with addition (ablation) | CameraAnything · Synthetic | PSNR: 15.62 -> 16.51 | Improved |
| use RoPE injection instead of AdaLN for Plucker rays (ablation) | CameraAnything · Synthetic | PSNR: 16.51 -> 16.66 | Improved |
| use RoPE instead of AdaLN (ablation) | CameraAnything · Synthetic | LPIPS: 0.368 -> 0.372 | Mixed |
| jointly train focal and resolution with extrinsics (ablation) | CameraAnything · Synthetic extrinsic tasks | PSNR: 14.00 -> 14.13 | No clear effect |
| add identity samples (decoupled control dimensions) (ablation) | CameraAnything · Synthetic focal tasks | PSNR: 15.94 -> 20.03 | Improved |
| add identity samples (decoupled control dimensions) (ablation) | CameraAnything · Synthetic resolution tasks | PSNR: 14.64 -> 20.05 | Improved |
| use CameraAnything vs ReCamMaster for camera accuracy | CameraAnything · DAVIS | RotErr: 5.12 -> 2.76 | Improved |
| use CameraAnything vs ReCamMaster on VBench | CameraAnything · Synthetic / DAVIS | VBench consistency and flicker: DAVIS Flicker 0.9596 vs 0.9463 | Hurt |
| use CameraAnything vs VACE on subject/background consistency | CameraAnything · Synthetic resolution editing | Background consistency: 0.9225 -> 0.9100 | Hurt |
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.
- ReCamMaster — ReCamMaster: Camera-Controlled Generative Rendering from A Single Video (ICCV 2025) · better on 7 of 8 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
- ReCamMaster: Camera-Controlled Generative Rendering from A Single Video (ICCV 2025 · similarity 0.87)
- OmniCamera: A Unified Framework for Multi-task Video Generation with Arbitrary Camera Control (ECCV 2026 · similarity 0.87)
- PoseAnything: General Pose-guided Video Generation with Part-aware Temporal Coherence (CVPR 2026 · similarity 0.87)
- VD3D: Taming Large Video Diffusion Transformers for 3D Camera Control (ICLR 2025 · similarity 0.86)
- AC3D: Analyzing and Improving 3D Camera Control in Video Diffusion Transformers (CVPR 2025 · similarity 0.85)
- CineOrchestra: Unified Entity-Centric Conditioning for Cinematic Video Generation (NeurIPS 2026 · similarity 0.85)
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.
