Research map / Video / Video segmentation and tracking
Analysis of Video Quality Datasets via Design of Minimalistic Video Quality Models
Wei Sun, Wen Wen, Xiongkuo Min, Long Lan, Guangtao Zhai, Kede Ma
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 motion features from the temporal quality analyzer (ablation) | Model (I) ResNet-50 · LBVD | SRCC: 0.704 -> 0.882 | Improved |
| add motion features from the temporal quality analyzer (ablation) | Model (I) ResNet-50 · LIVE-VQC | SRCC: 0.653 -> 0.752 | Improved |
| add motion features from the temporal quality analyzer (ablation) | Model (I) ResNet-50 · KoNViD-1k | SRCC: 0.764 -> 0.797 | Improved |
| add motion features from the temporal quality analyzer (ablation) | Model (I) ResNet-50 · CVD2014 | SRCC: 0.867 -> 0.852 | Hurt |
| add IQA pre-trained initialization (ablation) | Model (I) ResNet-50 · LIVE-Qualcomm | SRCC: 0.651 -> 0.762 | Improved |
| add IQA pre-trained initialization (ablation) | Model (I) ResNet-50 · LSVQ | SRCC: 0.833 -> 0.834 | No clear effect |
| replace ResNet-50 with Swin Transformer-B spatial analyzer (ablation) | Model (I) · CVD2014 | SRCC: 0.867 -> 0.833 | Hurt |
| scale ResNet-50 up to ResNet-200 (ablation) | Model (I) · LSVQ | SRCC: 0.819 -> 0.816 | No clear effect |
| scale Swin Transformer from Tiny to Base (ablation) | Model (I) · LSVQ | SRCC: 0.842 -> 0.852 | Improved |
| replace ImageNet-1k with CLIP initialization (ablation) | Model (I) ResNet-50 · LSVQ | SRCC: 0.819 -> 0.842 | Improved |
| replace ImageNet-1k with MAE initialization (ablation) | ViT-B · LSVQ | SRCC: 0.848 -> 0.846 | No clear effect |
| replace ImageNet-1k with ImageNet-22k initialization (ablation) | Swin Transformer-B · LSVQ | SRCC: 0.852 -> 0.857 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- FAST-VQA — FAST-VQA: Efficient End-to-End Video Quality Assessment with Fragment Sampling (ECCV 2022) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Modular Blind Video Quality Assessment (CVPR 2024 · similarity 0.93)
- PTM-VQA: Efficient Video Quality Assessment Leveraging Diverse PreTrained Models from the Wild (CVPR 2024 · similarity 0.90)
- KVQ: Boosting Video Quality Assessment via Saliency-guided Local Perception (CVPR 2025 · similarity 0.89)
- AIGV-Assessor: Benchmarking and Evaluating the Perceptual Quality of Text-to-Video Generation with LMM (CVPR 2025 · similarity 0.88)
- Scaling and Masking: A New Paradigm of Data Sampling for Image and Video Quality Assessment (AAAI 2024 · similarity 0.87)
- Learned Scanpaths Aid Blind Panoramic Video Quality Assessment (CVPR 2024 · 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.
