Research map / Video / Video segmentation and tracking
JSI-GAN: GAN-Based Joint Super-Resolution and Inverse Tone-Mapping with Pixel-Wise Task-Specific Filters for UHD HDR Video
Soo Ye Kim, Jihyong Oh, Munchurl Kim
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 the DR subnet to IR using division-based detail decomposition (ablation) | JSInet · JSInet subnet ablation | PSNR (dB): 0.44 dB gain | Improved |
| Add the LCE subnet to IR + DR (ablation) | JSInet · JSInet subnet ablation | PSNR (dB): 0.11 dB | Improved |
| Use division rather than subtraction to decompose the detail layer (ablation) | JSInet with DR and LCE · JSInet subnet ablation | PSNR (dB): 0.31 dB | Improved |
| Remove concatenation of DR intermediate features (ablation) | JSInet IR + DR · JSInet subnet ablation | PSNR (dB): 0.02 dB drop | Hurt |
| Concatenate LCE intermediate features with IR and DR features (ablation) | JSInet · JSInet subnet ablation | PSNR (dB): 0.25 dB drop | Hurt |
| Stack the input image with the detail layer as DR input (ablation) | JSInet IR + DR · JSInet subnet ablation | PSNR (dB) | No clear effect |
| Use location-varying filter generation for DR instead of direct detail reconstruction (ablation) | JSInet IR + DR · JSInet IR/DR filter ablation | PSNR (dB): 0.15 dB | Improved |
| Use filter generation for IR rather than direct reconstruction (ablation) | JSInet IR + DR · JSInet IR/DR filter ablation | PSNR (dB): 35.88 -> 34.14 | Hurt |
| Use 1D separable kernels instead of 2D local filters for LCE (ablation) | JSInet LCE subnet · JSInet filter ablation | PSNR (dB): 0.07 dB | Hurt |
| Add feature-matching loss to basic adversarial training (ablation) | JSI-GAN · 4K HDR dataset ×2 | PSNR (dB): 1.9dB | Improved |
| Add feature-matching loss to basic adversarial training (ablation) | JSI-GAN · 4K HDR dataset ×4 | PSNR (dB): 1.37dB | Improved |
| Add detail-component adversarial and feature-matching losses (ablation) | JSI-GAN with feature matching · 4K HDR dataset ×2 | PSNR (dB): 35.63 -> 35.73 | Improved |
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.
- Deep SR-ITM — Deep SR-ITM: Joint Learning of Super-Resolution and Inverse Tone-Mapping for 4K UHD HDR Applications (ICCV 2019) · better on 4 of 4 shared results
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
- Deep SR-ITM: Joint Learning of Super-Resolution and Inverse Tone-Mapping for 4K UHD HDR Applications (ICCV 2019 · similarity 0.90)
- SAVSR: Arbitrary-Scale Video Super-Resolution via a Learned Scale-Adaptive Network (AAAI 2024 · similarity 0.87)
- Noise-Started One-Step Image Super-Resolution via LR-Conditioned SplitMeanFlow and GAN Refinement (NeurIPS 2026 · similarity 0.87)
- Towards Real-World HDR Video Reconstruction: A Large-Scale Benchmark Dataset and A Two-Stage Alignment Network (CVPR 2024 · similarity 0.86)
- Zero-Shot Structure-Preserving Diffusion Model for High Dynamic Range Tone Mapping (CVPR 2024 · similarity 0.86)
- Uncertainty-Aware GAN for Single Image Super Resolution (AAAI 2024 · similarity 0.86)
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.
