Research map / Image generation / Super-resolution
OmniZoomer: Learning to Move and Zoom in on Sphere at High-Resolution
Zidong Cao, Hao Ai, Yan-Pei Cao, Ying Shan, Xiaohu Qie, Lin Wang
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 |
|---|---|---|---|
| apply Moebius transform at HR feature level instead of input image level (ablation) | OmniZoomer · ODI-SR x8 | WS-PSNR: 26.06 -> 27.48 | Improved |
| apply Moebius transform at HR feature level instead of HR image level (ablation) | OmniZoomer · ODI-SR x8 | WS-PSNR: 27.41 -> 27.48 | Improved |
| replace Jacobian resampling with spherical resampling (ablation) | OmniZoomer · ODI-SR x8 | WS-PSNR: 27.39 -> 27.46 | Improved |
| make Jacobian 2D rotation estimation learnable with an MLP (ablation) | Jacobian resampling · ODI-SR x8 | WS-PSNR: 27.39 -> 27.40 | No clear effect |
| add ResBlocks to the decoder (ablation) | OmniZoomer · ODI-SR x8 | WS-PSNR: 27.46 -> 27.48 | Improved |
| keep feature up-sampling and Moebius transform as separate processes rather than integrated (ablation) | OmniZoomer · ODI-SR x8 | WS-PSNR: 27.41 -> 27.48 | Improved |
| learn Moebius transform in network (OmniZoomer-RCAN) instead of SR then image-level transform (RCAN) | RCAN · ODIM SUN360 x8 | WS-PSNR: 27.04 -> 27.34 | Improved |
| use OmniZoomer for direct SR with identity transform | OmniZoomer-RCAN · ODI-SR x8 direct SR | WS-PSNR: 24.53 vs LAU-Net+ 24.63 | Mixed |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 2 on most shared results.
- LAU-Net+ — LAU-Net: Latitude Adaptive Upscaling Network for Omnidirectional Image Super-Resolution (CVPR 2021) · better on 1 of 2 shared results
- SphereSR — SphereSR: 360deg Image Super-Resolution With Arbitrary Projection via Continuous Spherical Image Representation (CVPR 2022) · better on 1 of 1 shared result
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
- ODGS: 3D Scene Reconstruction from Omnidirectional Images with 3D Gaussian Splattings (NeurIPS 2024 · similarity 0.85)
- OmniLocalRF: Omnidirectional Local Radiance Fields from Dynamic Videos (CVPR 2024 · similarity 0.85)
- ODI-Bench: Can MLLMs Understand Immersive Omnidirectional Environments? (ICLR 2026 · similarity 0.85)
- Spherical Pseudo-Cylindrical Representation for Omnidirectional Image Super-resolution (AAAI 2024 · similarity 0.84)
- SC-OmniGS: Self-Calibrating Omnidirectional Gaussian Splatting (ICLR 2025 · similarity 0.84)
- OmniSplat: Taming Feed-Forward 3D Gaussian Splatting for Omnidirectional Images with Editable Capabilities (CVPR 2025 · 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.
