Research map / Continual learning / Online and replay-based
Continual Learning for Blind Image Quality Assessment
Weixia Zhang, Dingquan Li, Chao Ma, Guangtao Zhai, Xiaokang Yang, 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 K-means gating at inference to MH-CL (ablation) | MH-CL · BIQA chronological | mSRCC: 0.7079 -> 0.7225 | Improved |
| add K-means gating to LwF (ablation) | LwF · BIQA chronological | mSRCC: 0.6693 -> 0.8150 | Improved |
| add K-means gating to EWC (ablation) | EWC · BIQA chronological | mSRCC: 0.6815 -> 0.7919 | Improved |
| add K-means gating to MAS (ablation) | MAS · BIQA chronological | mSRCC: 0.6170 -> 0.7795 | Improved |
| add K-means gating to SI (ablation) | SI · BIQA chronological | mSRCC: 0.7323 -> 0.7848 | Improved |
| replace KG with nearest-mean-of-exemplars gate (ablation) | LwF-KG · BIQA chronological | mPSI: 0.9180 -> 0.9046 | Hurt |
| replace KG with Expert Gate (ablation) | LwF-KG · BIQA chronological | mPSI: 0.9180 -> 0.9061 | Hurt |
| add KG during training (ablation) | LwF-KG · BIQA chronological | mPSI: 0.9180 -> 0.9175 | No clear effect |
| use two-stream backbone instead of single stream with bilinear pooling (ablation) | LwF-KG · BIQA chronological | mPSI: 0.9046 -> 0.9180 | Improved |
| use bilinear pooling instead of global average pooling (ablation) | single stream · BIQA chronological | mPSI: 0.8645 -> 0.9046 | Improved |
| use test-time task oracle instead of KG | LwF · BIQA chronological | mPSI: 0.9180 -> 0.9314 | Improved |
| reverse chronological task order (KADID-10K first) (ablation) | LwF-KG · BIQA task order | mPSI: 0.9180 -> 0.8984 | Hurt |
Closest papers by meaning
- Semi-Supervised Blind Image Quality Assessment through Knowledge Distillation and Incremental Learning (AAAI 2024 · similarity 0.91)
- Integrating Global Context Contrast and Local Sensitivity for Blind Image Quality Assessment (ICML 2024 · similarity 0.89)
- Feature Denoising Diffusion Model for Blind Image Quality Assessment (AAAI 2025 · similarity 0.89)
- Towards Syn-to-Real IQA: A Novel Perspective on Reshaping Synthetic Data Distributions (NeurIPS 2025 · similarity 0.88)
- Distilling Spatially-Heterogeneous Distortion Perception for Blind Image Quality Assessment (CVPR 2025 · similarity 0.87)
- Adaptive Image Quality Assessment via Teaching Large Multimodal Model to Compare (NeurIPS 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.
