Research map / Video / Video segmentation and tracking
Deep Contextual Video Compression
Jiahao Li, Bin Li, Yan Lu
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 |
|---|---|---|---|
| Use DCVC instead of x265 (veryslow) | Video compression · HEVC Class B | BD-Bitrate: -26.0% | Improved |
| Use DCVC instead of DVCPro | Video compression · MCL-JCV, UVG, and HEVC Classes B, C, D, E | BD-Bitrate: DCVC: -23.9%, -25.3%, -26.0%, -5.8%, -17.5%, -11.9%; DVCPro: -4.1%, -7.9%, -9.0%, 7.2%, -6.9%, 17.2% | Improved |
| Add feature-domain conditional concatenation rather than using RGB prediction concatenation (ablation) | DCVC · Conditional coding and temporal prior ablation | Bitrate increase: 5.4% -> 0.0% | Improved |
| Add conditional concatenation of context features (ablation) | DCVC without temporal prior · Conditional coding and temporal prior ablation | Bitrate increase: 12.9% -> 4.6% | Improved |
| Add temporal prior (ablation) | DCVC with conditional coding · Conditional coding and temporal prior ablation | Bitrate increase: 4.6% -> 0.0% | Improved |
| Add conditional concatenation and temporal prior (ablation) | DCVC · Conditional coding and temporal prior ablation | Bitrate increase: 12.9% -> 0.0% | Improved |
| Add temporal prior while retaining spatial prior (ablation) | DCVC entropy model · Entropy model ablation | Bitrate increase: 4.6% -> 0.0% | Improved |
| Add temporal prior without spatial prior (ablation) | Hyper-prior entropy model · Entropy model ablation | Bitrate increase: 60.9% -> 3.8% | Improved |
| Add spatial prior while retaining temporal prior (ablation) | DCVC entropy model · Entropy model ablation | Bitrate increase: 3.8% -> 0.0% | Improved |
| Remove spatial prior (ablation) | DCVC entropy model · Entropy model ablation | Bitrate increase: 0.0% -> 3.8% | Hurt |
| Remove temporal prior (ablation) | DCVC entropy model · Entropy model ablation | Bitrate increase: 0.0% -> 4.6% | Hurt |
| Remove both spatial and temporal priors (ablation) | DCVC entropy model · Entropy model ablation | Bitrate increase: 0.0% -> 60.9% | Hurt |
Closest papers by meaning
- Scene Matters: Model-based Deep Video Compression (ICCV 2023 · similarity 0.89)
- Neural Video Compression with Context Modulation (CVPR 2025 · similarity 0.89)
- Conditional Entropy Coding for Efficient Video Compression (ECCV 2020 · similarity 0.89)
- Neural Video Compression with Feature Modulation (CVPR 2024 · similarity 0.88)
- An End-to-End Learning Framework for Video Compression (TPAMI 2020 · similarity 0.87)
- Neural Video Compression With Diverse Contexts (CVPR 2023 · similarity 0.87)
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.
