Research map / Video / Video segmentation and tracking
DVC: An End-To-End Deep Video Compression Framework
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 |
|---|---|---|---|
| Jointly optimize motion estimation for the rate-distortion objective instead of fixing the pretrained flow model (ablation) | DVC · Optical flow, warped frame | Bpp: 0.044 -> 0.029 | Helped |
| Jointly optimize motion estimation for the rate-distortion objective instead of fixing the pretrained flow model (ablation) | DVC · Optical flow, warped frame | PSNR: 27.33 -> 28.17 | Helped |
| Add the motion compensation refinement network (ablation) | DVC · Video compression, same Bpp | PSNR: about 1.0 dB | Helped |
| Use buffered reconstructed reference frames instead of original frames during training (ablation) | DVC · Video compression, same Bpp | PSNR: about 0.2dB | Helped |
| Add the motion-vector encoder-decoder network to compress optical flow (ablation) | DVC · Video compression, same Bpp | PSNR: more than 2 dB | Helped |
| Add the motion-vector encoder-decoder network instead of directly encoding raw optical flow (ablation) | DVC · Optical flow, warped frame | Bpp: 0.20Bpp -> 0.029Bpp | Helped |
| Add motion information rather than coding each frame independently (ablation) | DVC · Video compression | PSNR: more than 2dB | Helped |
| Use DVC instead of H.264 | DVC · HEVC sequences, average over all sequences | BDBR (PSNR, %): -19.22 | Helped |
| Use DVC instead of H.264 | DVC · HEVC sequences, average over all sequences | BDBR (MS-SSIM, %): -29.32 | Helped |
| Use DVC instead of H.265 | DVC · HEVC sequences, average over all sequences | BDBR (MS-SSIM, %): -29.32 vs -21.73 | Helped |
| Use DVC instead of H.265 | DVC · HEVC sequences, average over all sequences | BDBR (PSNR, %): -19.22 vs -25.06 | Hurt |
| Use DVC instead of Wu_ECCV2018 | DVC · UVG | PSNR: about 0.6dB | Helped |
Closest papers by meaning
- Offline and Online Optical Flow Enhancement for Deep Video Compression (AAAI 2024 · similarity 0.89)
- Towards Practical Real-Time Neural Video Compression (CVPR 2025 · similarity 0.88)
- Neural Video Compression with Feature Modulation (CVPR 2024 · similarity 0.86)
- Neural Video Compression with Context Modulation (CVPR 2025 · similarity 0.86)
- PNVC: Towards Practical INR-based Video Compression (AAAI 2025 · similarity 0.86)
- Task-Aware Encoder Control for Deep Video Compression (CVPR 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.