Research map / 3D vision / Depth and stereo
Fast Light Field Reconstruction With Deep Coarse-To-Fine Modeling of Spatial-Angular Clues
Henry Wing Fung Yeung, Junhui Hou, Jie Chen, Yuk Ying Chung, Xiaoming Chen
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 extrapolated input view positions (1 row and column) together with interpolation (ablation) | Ours 8L · 2x2-8x8, 222 real-world LFs | PSNR: 37.55 -> 39.93 | Improved |
| extrapolate 2 rows and columns of input views (ablation) | Ours 8L · 2x2-8x8, 222 real-world LFs | PSNR: 37.55 -> 40.09 | Improved |
| extrapolate 2 instead of 1 row/column of input views (ablation) | Ours Extra. 1 · 2x2-8x8, 222 real-world LFs | PSNR: 39.93 -> 40.09 | Improved |
| use 16 alternating convolutions in synthesis network vs 8 (ablation) | Ours · 2x2-8x8, 222 real-world LFs | PSNR: 37.55 -> 37.76 | Improved |
| use proposed model instead of Kalantari et al. | Ours 4L · 3x3-7x7 average over datasets | PSNR: 43.18 -> 44.06 | Improved |
| use proposed model instead of Wu et al. | Ours 4L · 3x3-7x7 average over datasets | PSNR: 40.75 -> 44.06 | Improved |
| use proposed model on microscope Neurons 20x | Ours 4L vs Kalantari et al. · Neurons 20x | PSNR: 28.45 -> 30.69 | Improved |
| use proposed Ours 8L instead of Wu et al. | Ours 8L · 3x3-9x9 HCI Buddha and Mona | PSNR: 43.79 -> 44.72 | Improved |
| use proposed 16L model instead of Kalantari et al. | Ours 16L · 2x2-8x8 Occlusions | PSNR: 31.81 -> 32.68 | Improved |
| use proposed model instead of Wu et al. on SSIM | Ours 4L · 3x3-7x7 average | SSIM: 0.9861 -> 0.9889 | Improved |
| use proposed model instead of Wu et al. (HCI, SSIM-based comparison) | Ours 8L · 3x3-9x9 HCI Buddha and Mona | SSIM: 0.9981 -> 0.9896 | Hurt |
Closest papers by meaning
- Deep Coarse-to-Fine Dense Light Field Reconstruction With Flexible Sampling and Geometry-Aware Fusion (TPAMI 2020 · similarity 0.95)
- Deep Spatial-angular Regularization for Compressive Light Field Reconstruction over Coded Apertures (ECCV 2020 · similarity 0.92)
- Deep Spatial-Angular Regularization for Light Field Imaging, Denoising, and Super-Resolution (TPAMI 2021 · similarity 0.91)
- Light Field Reconstruction via Deep Adaptive Fusion of Hybrid Lenses (TPAMI 2023 · similarity 0.91)
- Learning Light Field Angular Super-Resolution via a Geometry-Aware Network (AAAI 2020 · similarity 0.90)
- High-Dimensional Dense Residual Convolutional Neural Network for Light Field Reconstruction (TPAMI 2019 · similarity 0.90)
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.
