Research map / 3D vision / 3D generation
EpipolarGAN: Omnidirectional Image Synthesis with Explicit Camera Control
Christopher May, Daniel Aliaga
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 epipolar loss on intermediate features (ablation) | EpipolarGAN (config C) · Pano3D 2000 kimgs | SfM RMSE x10^-3: 7.36 -> 7.25 | Improved |
| add epipolar loss on intermediate features (ablation) | EpipolarGAN (config C) · Pano3D 2000 kimgs | SfM NP: 0.75 -> 0.775 | Improved |
| add polar filtering (per-row filters) (ablation) | EpipolarGAN (config B) · Pano3D 2000 kimgs | SfM RMSE x10^-3: 7.47 -> 7.36 | No clear effect |
| add polar filtering (ablation) | EpipolarGAN (config B) · Pano3D 2000 kimgs | polar distortion (visual) | Improved |
| full pipeline vs StyleGAN3-T at 2000 kimgs (adds camera control) (ablation) | StyleGAN3-T · Pano3D | FID: 16.7 -> 15.0 | Improved |
| full pipeline vs StyleGAN3-T at 2000 kimgs (ablation) | StyleGAN3-T · Pano3D | FID-Poles: 73.0 -> 58.6 | Improved |
| use blended polar filter instead of per-row filters (ablation) | EpipolarGAN polar filtering · training speed | s/kimg: 983 -> 236 s/kimg | Improved |
| EpipolarGAN vs GSN | omnidirectional GAN · Pano3D vs GSN pretrained | SfM NP: 0.77 -> 0.716 | 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.
- GAUDI — GAUDI: A Neural Architect for Immersive 3D Scene Generation (NeurIPS 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
- Towards Unsupervised Learning of Generative Models for 3D Controllable Image Synthesis (CVPR 2020 · similarity 0.88)
- EpiGRAF: Rethinking training of 3D GANs (NeurIPS 2022 · similarity 0.87)
- EscherNet: A Generative Model for Scalable View Synthesis (CVPR 2024 · similarity 0.87)
- Epipolar-Free 3D Gaussian Splatting for Generalizable Novel View Synthesis (NeurIPS 2024 · similarity 0.87)
- EpiDiff: Enhancing Multi-View Synthesis via Localized Epipolar-Constrained Diffusion (CVPR 2024 · similarity 0.86)
- ERUPT: Efficient Rendering with Unposed Patch Transformer (CVPR 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.
