Research map / Video / Video segmentation and tracking
Video compression dataset and benchmark of learning-based video-quality metrics
Anastasia Antsiferova, Sergey Lavrushkin, Maksim Smirnov, Aleksandr Gushchin, Dmitriy S. Vatolin, Dmitriy Kulikov
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 |
|---|---|---|---|
| include chroma (YUV) components in VMAF instead of luma only (ablation) | VMAF · full dataset | SROCC: 0.942 -> 0.943 | Improved |
| use VMAF NEG (no enhancement gain) instead of original VMAF (ablation) | VMAF · full dataset | SROCC: 0.943 -> 0.917 | Hurt |
| use no-reference MDTVSFA instead of best full-reference VMAF | VMAF · full dataset | SROCC: 0.943 vs 0.929 | No clear effect |
| use mixed-dataset training MDTVSFA instead of VSFA | VSFA · full dataset | SROCC: 0.905 -> 0.929 | Improved |
| evaluate at high bitrate (>6,000 kbps) vs all videos | Y-VMAF (v061) · video quality metrics | SROCC: 0.942 -> 0.453 | Hurt |
| evaluate on VVC-encoded videos vs best on all | MDTVSFA · video quality metrics | SROCC: 0.929 -> 0.882 | Hurt |
| use learned perceptual metrics (LPIPS, DISTS) vs classical SSIM/PSNR | LPIPS · full dataset | SROCC: 0.906 (YUV-SSIM) -> 0.749 (LPIPS) | Hurt |
Closest papers by meaning
- Perceptual Video Compression with Neural Wrapping (CVPR 2025 · similarity 0.85)
- Modular Blind Video Quality Assessment (CVPR 2024 · similarity 0.83)
- Neural Video Compression with Feature Modulation (CVPR 2024 · similarity 0.83)
- PTM-VQA: Efficient Video Quality Assessment Leveraging Diverse PreTrained Models from the Wild (CVPR 2024 · similarity 0.83)
- NVRC: Neural Video Representation Compression (NeurIPS 2024 · similarity 0.83)
- High Dynamic Range Video Compression: A Large-Scale Benchmark Dataset and A Learned Bit-depth Scalable Compression Algorithm (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.
