Research map / Video / Video segmentation and tracking
Learn to Recover Visible Color for Video Surveillance in a Day
Guangming Wu, Yinqiang Zheng, Zhiling Guo, Zekun Cai, Xiaodan Shi, Xin Ding, Yifei Huang, Yimin Guo, Ryosuke Shibasaki
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 state synchronization module (ablation) | SSN · VSIAD ablation | LPIPS: 0.212 -> 0.179 | Improved |
| add state synchronization module (ablation) | SSN · VSIAD ablation | PSNR: 21.419 -> 22.013 | Improved |
| use RGB input instead of grayscale (ablation) | SSN · VSIAD ablation | LPIPS: 14.3% LPIPS loss; 0.209 -> 0.179 | Improved |
| use RGB input instead of grayscale (ablation) | SSN · VSIAD ablation | PSNR: 21.535 -> 22.013 | Improved |
| use DSSIM instead of L1 for texture loss (ablation) | SSN · VSIAD ablation | LPIPS: 0.188 -> 0.179 | Improved |
| use DSSIM instead of L1 for texture loss (ablation) | SSN · VSIAD ablation | PSNR: 21.910 -> 22.013 | Improved |
| use instance normalization instead of group normalization (ablation) | SSN · VSIAD ablation | LPIPS: 0.182 -> 0.179 | Improved |
| use instance normalization instead of group normalization (ablation) | SSN · VSIAD ablation | PSNR: 23.145 -> 22.013 | Hurt |
| fine-tune Iizuka et al. on VSIAD visible images | Iizuka et al. · VSIAD | PSNR: +-0.2 PSNR | No clear effect |
| use SSN instead of pix2pixHD | SSN vs pix2pixHD · VSIAD VNIR2VC | PSNR: 25.003 -> 24.336 | Hurt |
| use SSN instead of pix2pixHD | SSN vs pix2pixHD · VSIAD VNIR2VC | LPIPS: 0.139 -> 0.109 | Improved |
| use SSN instead of pix2pixHD | SSN vs pix2pixHD · VSIAD NIR2VC | PSNR: 19.654 -> 19.690 | No clear effect |
Closest papers by meaning
- Optimal LED Spectral Multiplexing for NIR2RGB Translation (CVPR 2022 · similarity 0.87)
- An Integrated Enhancement Solution for 24-Hour Colorful Imaging (AAAI 2020 · similarity 0.87)
- NIR-assisted Video Enhancement via Unpaired 24-hour Data (ICCV 2023 · similarity 0.87)
- Learning to See Moving Objects in the Dark (ICCV 2019 · similarity 0.86)
- NightCC: Nighttime Color Constancy via Adaptive Channel Masking (CVPR 2024 · similarity 0.83)
- Seeing Dark Videos via Self-Learned Bottleneck Neural Representation (AAAI 2024 · similarity 0.82)
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.
