Research map / 3D vision / 3D generation
PanoFree: Tuning-Free Holistic Multi-view Image Generation with Cross-view Self-Guidance
Aoming Liu, Zhong Li, Zhang Chen, Nannan Li, Yi Xu, Bryan Plummer
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 cross-view guidance (SDEdit with earlier views) (ablation) | PanoFree sequential warp-and-inpaint · Planar Panorama (ablation) | FID: 24.75 -> 20.63 | Improved |
| add cross-view guidance (ablation) | PanoFree sequential warp-and-inpaint · 360 Panorama (ablation) | Intra-LPIPS: 71.34 -> 69.21 | Improved |
| add distance-based risky area erasing (ablation) | PanoFree with cross-view guidance · Planar Panorama (ablation) | FID: 20.63 -> 18.21 | Improved |
| add color and smoothness based erasing (ablation) | PanoFree with distance/edge erasing · 360 Panorama (ablation) | FID: 26.69 -> 26.51 | No clear effect |
| use PanoFree instead of vanilla sequential generation | Stable Diffusion v2.0 inpainting · 360 Panorama | FID: 32.28 -> 25.84 | Improved |
| use PanoFree instead of MVDiffusion (fine-tuned) | 360 panorama generation · 360 Panorama | FID: 37.89 -> 25.84 | Improved |
| use PanoFree instead of SyncDiffusion | Planar panorama generation · Planar Panorama | Intra-LPIPS: 64.48 -> 65.34 | Hurt |
| use PanoFree instead of SyncDiffusion | Planar panorama generation · Planar Panorama | Time (s): 128 -> 26 | Improved |
| use PanoFree instead of MultiDiffusion | Planar panorama generation · Planar Panorama | Cross-LPIPS: 69.92 -> 69.68 | Improved |
| use PanoFree instead of SyncDiffusion | Planar panorama generation · Planar user study | Diversity preference (%): 35.3 -> 64.7 | Improved |
| use large guidance scale during expansion and closing stages (ablation) | Full spherical panorama generation · Full Spherical Panorama | artifacts (qualitative) | Hurt |
Closest papers by meaning
- HORIZON: High-Resolution Semantically Controlled Panorama Synthesis (AAAI 2024 · similarity 0.87)
- ViewPoint: Panoramic Video Generation with Pretrained Diffusion Models (NeurIPS 2025 · similarity 0.87)
- DiffPano: Scalable and Consistent Text to Panorama Generation with Spherical Epipolar-Aware Diffusion (NeurIPS 2024 · similarity 0.87)
- Taming Stable Diffusion for Text to 360 Panorama Image Generation (CVPR 2024 · similarity 0.87)
- 4K4DGen: Panoramic 4D Generation at 4K Resolution (ICLR 2025 · similarity 0.87)
- PanoWan: Lifting Diffusion Video Generation Models to 360$^\circ$ with Latitude/Longitude-aware Mechanisms (NeurIPS 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.
