Research map / Video / Video segmentation and tracking
Binarized High-Efficiency RAW Video Restoration and Beyond
Tianyu Zhu, Ying Fu, Hesong Li, Gengchen Zhang, Xin Yuan, Yulun Zhang
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 temporal information interaction to the BIIM decoder (ablation) | BinRVR · LLRVD | PSNR: 37.00 -> 37.21 | Improved |
| add temporal information interaction to the BIIM encoder (ablation) | BinRVR · LLRVD | PSNR: 37.06 -> 37.21 | Improved |
| add spatial information interaction to the decoder (ablation) | BinRVR · LLRVD | PSNR: 37.04 -> 37.21 | Improved |
| use full TII in encoder as in decoder (ablation) | BinRVR · LLRVD | PSNR: 37.21 -> 37.15 | Hurt |
| use simplified encoder-style TII in decoder (ablation) | BinRVR · LLRVD | PSNR: 37.21 -> 37.14 | Hurt |
| change strip kernel from 1x11 to 1x13 (ablation) | BinRVR · LLRVD | PSNR: 37.21 -> 37.17 | No clear effect |
| change strip kernel from 1x11 to 1x9 (ablation) | BinRVR · LLRVD | PSNR: 37.21 -> 37.16 | No clear effect |
| change branch ratio to 1/4 (ablation) | BinRVR · LLRVD | PSNR: 37.21 -> 37.17 | No clear effect |
| use hybrid sliding-window recurrent instead of pure sliding window (ablation) | BinRVR · LLRVD | PSNR: 36.83 -> 37.21 | Improved |
| use hybrid sliding-window recurrent instead of pure recurrent (ablation) | BinRVR · LLRVD | PSNR: 36.98 -> 37.21 | Improved |
| increase recurrent stride from 1 to 2 (ablation) | BinRVR · LLRVD | PSNR: 0.40 dB drop in PSNR, nearly 40% FLOPs reduced | Hurt |
| replace RGB2RGB pipeline with RAW2RAW+ISP | low-light enhancement · LLRVD (RGB-domain eval) | PSNR: 24.91 -> 30.59 | Improved |
Closest papers by meaning
- Binarized Low-light Raw Video Enhancement (CVPR 2024 · similarity 0.91)
- BiViT: Extremely Compressed Binary Vision Transformers (ICCV 2023 · similarity 0.87)
- Basic Binary Convolution Unit for Binarized Image Restoration Network (ICLR 2023 · similarity 0.86)
- Flexible Residual Binarization for Image Super-Resolution (ICML 2024 · similarity 0.86)
- DeNC: Unleash Neural Codecs in Video Streaming with Diffusion Enhancement (AAAI 2025 · similarity 0.84)
- SeedVR: Seeding Infinity in Diffusion Transformer Towards Generic Video Restoration (CVPR 2025 · similarity 0.84)
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.
