Research map / Video / Video generation
Captain Safari: A World Engine with Pose-Aligned 3D Memory
Yu-Cheng Chou, Xingrui Wang, Yitong Li, Jiahao Wang, Hanting Liu, Cihang Xie, Alan Yuille, Junfei Xiao
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 pose-aligned memory retrieval (ablation) | Captain Safari · OpenSafari test | MEt3R: 0.3720 -> 0.3690 | Improved |
| add pose-aligned memory retrieval (ablation) | Captain Safari · OpenSafari test | reconstruction rate: 0.912 -> 0.968 | Improved |
| add pose-aligned memory retrieval (ablation) | Captain Safari · OpenSafari test | AUC@30: 0.193 -> 0.200 | Improved |
| add pose-aligned memory retrieval (ablation) | Captain Safari · OpenSafari test | CosSim: 0.508 -> 0.563 | Improved |
| add pose-aligned memory retrieval (ablation) | Captain Safari · OpenSafari test | AUC@15: 0.068 -> 0.068 | No clear effect |
| add pose-aligned memory retrieval (ablation) | Captain Safari · OpenSafari test | FVD: 998.47 -> 1023.46 | Hurt |
| add pose-aligned memory retrieval (ablation) | Captain Safari · OpenSafari test | LPIPS: 0.504 -> 0.512 | Hurt |
| add pose-aligned memory retrieval (ablation) | Captain Safari · Human study | preference (average): 23.07% -> 67.33% | Improved |
| Captain Safari versus Wan2.2-5B-Control-Camera baseline | Wan2.2-Fun-5B-Control-Camera · OpenSafari test | FVD: 1387.75 -> 1023.46 | Improved |
| Captain Safari versus Real-CamI2V | Real-CamI2V · OpenSafari test | MEt3R: 0.3703 -> 0.3690 | 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.
- Real-CamI2V — RealCam-I2V: Real-World Image-to-Video Generation with Interactive Complex Camera Control (ICCV 2025) · better on 8 of 8 shared results
- Geometry Forcing — Geometry Forcing: Marrying Video Diffusion and 3D Representation for Consistent World Modeling (ICLR 2026) · better on 7 of 7 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
- WorldMem: Long-term Consistent World Simulation with Memory (NeurIPS 2025 · similarity 0.86)
- World Motion Models: Flexible Sequence Modeling of SE(3) Trajectories (NeurIPS 2026 · similarity 0.86)
- Video World Models with Long-term Spatial Memory (NeurIPS 2025 · similarity 0.86)
- WorldVLN: Autoregressive World Action Model for Aerial Vision-Language Navigation (NeurIPS 2026 · similarity 0.85)
- SyncWorld: Visual Calibration Enables World Models as Zero-Shot Simulators (NeurIPS 2026 · similarity 0.85)
- Learning 3D Persistent Embodied World Models (NeurIPS 2025 · 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.
