Research map / Image generation / Super-resolution
Disentangling Light Fields for Super-Resolution and Disparity Estimation
Yingqian Wang, Longguang Wang, Gaochang Wu, Jungang Yang, Wei An, Jingyi Yu A, Yulan Guo
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 angular branch to the spatial-only network (ablation) | DistgSSR · LF spatial SR 2x (average) | PSNR: 36.13 -> 38.62 | Improved |
| add EPI branches (ablation) | DistgSSR · LF spatial SR 2x (average) | PSNR: 38.62 -> 38.92 | Improved |
| add EPI branches (ablation) | DistgSSR · STFgantry 2x | PSNR: 0.41 dB | Improved |
| share weights between horizontal and vertical EPI branches (ablation) | DistgSSR · LF spatial SR 2x (average) | PSNR: 38.76 -> 38.92 | Improved |
| use Distg-Block instead of 4D convolutions (ablation) | DistgSSR · LF spatial SR 2x (average) | PSNR: 38.61 -> 38.92 | Improved |
| increase angular resolution from 2x2 to 9x9 (ablation) | DistgSSR · LF spatial SR 2x (average) | PSNR: 37.74 -> 39.18 | Improved |
| increase angular resolution from 7x7 to 9x9 (ablation) | DistgSSR · LF spatial SR 2x (average) | PSNR: 39.09 -> 39.18 | No clear effect |
| halve feature depth to 32 (DistgSSR_32 vs _64) (ablation) | DistgSSR · LF spatial SR 2x (average) | PSNR: 38.92 -> 38.33 | Hurt |
| add angular branch to the spatial-only network (ablation) | DistgASR · Angular SR 2x2->7x7 (average) | PSNR: 37.59 -> 39.89 | Improved |
| add spatial branch to the angular-only network (ablation) | DistgASR · Angular SR 2x2->7x7 (average) | PSNR: 38.42 -> 40.75 | Improved |
| add EPI branches (ablation) | DistgASR · Angular SR 2x2->7x7 (average) | PSNR: 39.89 -> 40.75 | Improved |
| share weights between horizontal and vertical EPI branches (ablation) | DistgASR · Angular SR 2x2->7x7 (average) | PSNR: 40.60 -> 40.75 | Improved |
Closest papers by meaning
- Light Field Super-Resolution using a Low-Rank Prior and Deep Convolutional Neural Networks (TPAMI 2019 · similarity 0.91)
- Residual Networks for Light Field Image Super-Resolution (CVPR 2019 · similarity 0.90)
- High-Dimensional Dense Residual Convolutional Neural Network for Light Field Reconstruction (TPAMI 2019 · similarity 0.89)
- Light Field Spatial Super-Resolution via Deep Combinatorial Geometry Embedding and Structural Consistency Regularization (CVPR 2020 · similarity 0.89)
- Learning Non-Local Spatial-Angular Correlation for Light Field Image Super-Resolution (ICCV 2023 · similarity 0.89)
- Learning Light Field Angular Super-Resolution via a Geometry-Aware Network (AAAI 2020 · similarity 0.89)
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.
