Research map / Video / Video generation
360Anything: Geometry-Free Lifting of Images and Videos to 360°
Ziyi Wu, Daniel Watson, Andrea Tagliasacchi, David Fleet, Marcus Brubaker, Saurabh Saxena
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 Circular Latent Encoding instead of vanilla VAE encoding (ablation) | 360Anything image · Seam discontinuity (image) | discontinuity score: 9.92 -> 3.87 | Improved |
| add Circular Latent Encoding instead of Blended Decoding (ablation) | 360Anything image · Seam discontinuity (image) | discontinuity score: 5.29 -> 3.87 | Improved |
| add Circular Latent Encoding instead of vanilla VAE encoding (ablation) | 360Anything video · Seam discontinuity (video) | discontinuity score: 35.52 -> 13.28 | Improved |
| add Circular Latent Encoding instead of Blended Decoding (ablation) | 360Anything video · Seam discontinuity (video) | discontinuity score: 19.84 -> 13.28 | Improved |
| add random camera (FoV, pitch, roll) augmentation in training (ablation) | 360Anything image · Laval Indoor FID | FID: 8.4 -> 8.0 | Improved |
| add random camera augmentation in training (ablation) | 360Anything image · Laval Indoor FAED | FAED: 10.4 -> 9.8 | Improved |
| add camera augmentation, tested at tilted conditioning views (ablation) | 360Anything image · Laval Indoor FID (average over conditioning views) | FID degradation: +6.48 -> +0.98 | Improved |
| use sequence concatenation instead of channel concatenation with GT camera (ablation) | 360Anything image · Laval Indoor FID (average over conditioning views) | FID degradation: +1.17 -> +0.98 | No clear effect |
| add canonical (gravity-aligned) training instead of non-canonical (ablation) | 360Anything video · Argus test videos (real trajectory) | FVD: 559.5 -> 470.8 | Improved |
| add canonical training instead of non-canonical (ablation) | 360Anything video · Argus test videos (real trajectory) | PSNR: 26.56 -> 24.02 | Hurt |
| increase conditioning FoV | 360Anything image · Laval Indoor | all metrics | Improved |
| tilt conditioning pitch and roll away from zero | 360Anything image · Laval Indoor | all metrics: degradations less than 1.0 | 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.
- ViewPoint* — ViewPoint: Panoramic Video Generation with Pretrained Diffusion Models (NeurIPS 2025) · better on 4 of 4 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
- Imagine360: Immersive 360 Video Generation from Perspective Anchor (NeurIPS 2025 · similarity 0.89)
- ViewPoint: Panoramic Video Generation with Pretrained Diffusion Models (NeurIPS 2025 · similarity 0.88)
- PanoWorld: Geometry-Consistent Panoramic Video World Modeling (NeurIPS 2026 · similarity 0.87)
- DepthMaster: Unified Monocular Depth Estimation for Perspective and Panoramic Images (NeurIPS 2026 · similarity 0.87)
- From an Image to a Scene: Learning to Imagine the World from a Million 360° Videos (NeurIPS 2024 · similarity 0.87)
- Splatter-360: Generalizable 360 Gaussian Splatting for Wide-baseline Panoramic Images (CVPR 2025 · similarity 0.86)
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.
