Research map / Image generation / Text-to-image diffusion
Spherical Image Generation From a Few Normal-Field-of-View Images by Considering Scene Symmetry
Takayuki Hara, Yusuke Mukuta, Tatsuya Harada
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 proposed symmetry-aware CVAE instead of PICNet | spherical image generation · Spherical images | FID: 39.6 -> 28.6 | Improved |
| use proposed method instead of Pix2pix | spherical image generation · Spherical images | FID: 46.3 -> 28.6 | Improved |
| widen input FOV from 30 to 120 degrees | proposed method · Front view | FID: 75.3 -> 27.4 | Improved |
| use top or bottom view instead of front 90 deg view | proposed method · Spherical images | FID: 30.4 -> 81.7 | Hurt |
| add L_gen to L_rec (ablation) | proposed method · Spherical images | FID: 41.8 -> 28.6 | Improved |
| add L_rec to L_gen (ablation) | proposed method · Spherical images | FID: 31.1 -> 28.6 | Improved |
| add L_gen to L_rec (ablation) | proposed method · Spherical images | L1 loss: 0.263 -> 0.279 | Hurt |
| add symmetry control function (ablation) | proposed method · Spherical images | FID: 40.3 -> 28.6 | Improved |
| add partial-image region weight w_kappa (ablation) | proposed method · Spherical images | FID: 37.2 -> 28.6 | Improved |
| add circular padding (ablation) | proposed method · Spherical images | seam discontinuity | Improved |
Closest papers by meaning
- Spherical Image Generation from a Single Image by Considering Scene Symmetry (AAAI 2021 · similarity 0.98)
- SphereDiffusion: Spherical Geometry-Aware Distortion Resilient Diffusion Model (AAAI 2024 · similarity 0.86)
- Symmetry Strikes Back: From Single-Image Symmetry Detection to 3D Generation (CVPR 2025 · similarity 0.85)
- Spherical Manifold Guided Diffusion Model for Panoramic Image Generation (CVPR 2025 · similarity 0.84)
- BerfScene: Bev-conditioned Equivariant Radiance Fields for Infinite 3D Scene Generation (CVPR 2024 · similarity 0.84)
- Autoregressive Omni-Aware Outpainting for Open-Vocabulary 360-Degree Image Generation (AAAI 2024 · similarity 0.84)
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.
