Research map / Video / Video segmentation and tracking
Coarse-To-Fine Deep Video Coding With Hyperprior-Guided Mode Prediction
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Replace single-scale compensation with coarse-to-fine motion compensation (ablation) | FVC(re-imp) · HEVC Class E | PSNR: +0.3dB at 0.08bpp | Helped |
| Add hyperprior-guided adaptive motion compression (HAMC) (ablation) | C2F · HEVC Class E | PSNR: +0.6dB at 0.03bpp | Helped |
| Add hyperprior-guided adaptive residual compression (HARC) along with C2F and HAMC (ablation) | C2F+HAMC · HEVC Class E | PSNR: outperforms the baseline FVC(re-imp) by 1.2dB at 0.03bpp | Helped |
| Use C2F motion compensation instead of single-scale compensation (ablation) | FVC(re-imp) motion compensation · HEVC Class B motion compensation visualization | PSNR: improved by 1.7dB with similar bit-rate cost | Helped |
| Use C2F motion compensation (ablation) | FVC(re-imp) · HEVC Class E motion compensation | PSNR: improved by 1.14dB, while costing similar bpps | Helped |
| Add HAMC and HARC to DVC (ablation) | DVC · MCL-JCV | PSNR: 0.5dB improvement at 0.1bpp | Helped |
| Use the complete C2F+HAMC+HARC framework instead of basic C2F (ablation) | C2F · 1920×1080 inference | speed: 3.43fps -> 3.41fps | No clear effect |
| Use the proposed method instead of H.265(HM) | Video compression · Average of HEVC | BDBR (%): 18.58 -> -7.39 | Helped |
| Use the proposed method instead of H.265(HM) | Video compression · HEVC Class C | BD-PSNR (dB): Ours -0.42 | Hurt |
| Use the proposed method instead of H.265(HM) | Video compression · HEVC Class B, Class C, Class D, Class E, UVG and MCL-JCV | BD-PSNR (dB): HEVC Class B 0.37; HEVC Class C -0.42; HEVC Class D 0.04; HEVC Class E 0.25; UVG 0.21; MCL-JCV 0.09 | Mixed |
Closest papers by meaning
- MMVC: Learned Multi-Mode Video Compression With Block-Based Prediction Mode Selection and Density-Adaptive Entropy Coding (CVPR 2023 · similarity 0.87)
- High Dynamic Range Video Compression: A Large-Scale Benchmark Dataset and A Learned Bit-depth Scalable Compression Algorithm (CVPR 2025 · similarity 0.85)
- Task-Aware Encoder Control for Deep Video Compression (CVPR 2024 · similarity 0.84)
- Towards Practical Real-Time Neural Video Compression (CVPR 2025 · similarity 0.84)
- Offline and Online Optical Flow Enhancement for Deep Video Compression (AAAI 2024 · similarity 0.84)
- Neural Video Compression with Context Modulation (CVPR 2025 · similarity 0.83)
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.