Research map / Image generation / Autoregressive and latent generation
Image Quality Assessment: Unifying Structure and Texture Similarity
Keyan Ding, Kede Ma, Shiqi Wang, Eero P. Simoncelli
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 |
|---|---|---|---|
| replace max pooling with l2 pooling (ablation) | LPIPS · SynTEX | SRCC: 0.589 -> 0.592 | Improved |
| add input image as zeroth stage (ablation) | LPIPS with l2 pooling · LIVE | SRCC: 0.938 -> 0.935 | No clear effect |
| replace Euclidean distance with global SSIM-like measures (ablation) | LPIPS + input image · TQD | SRCC: 0.409 -> 0.795 | Improved |
| replace Euclidean distance with global SSIM-like measures (ablation) | LPIPS + input image · LIVE_Aug | SRCC: 0.789 -> 0.881 | Improved |
| use local 11x11 SSIM instead of global SSIM (ablation) | LPIPS + input image · SynTEX | SRCC: 0.877 -> 0.744 | Hurt |
| local SSIM instead of Euclidean distance (ablation) | LPIPS + input image · TQD | SRCC: 0.409 -> 0.667 | Improved |
| add E2 texture-invariance training term (ablation) | global SSIM model · TQD | SRCC: 0.795 -> 0.910 | Improved |
| add E2 texture-invariance training term (ablation) | global SSIM model · LIVE_Aug | SRCC: 0.881 -> 0.928 | Improved |
| add E2 texture-invariance training term (ablation) | global SSIM model · TID2013 | SRCC: 0.835 -> 0.830 | Hurt |
| add E2 texture-invariance training term (ablation) | global SSIM model · LIVE | SRCC: 0.957 -> 0.954 | Hurt |
| add E2 term to model without SSIM distance (ablation) | LPIPS + input image · TID2013 | SRCC: 0.825 -> 0.776 | Hurt |
| DISTS vs LPIPS trained on BAPPS | DISTS · BAPPS | 2AFC all: 0.689 vs 0.692 | No clear effect |
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.
- PieAPP — PieAPP: Perceptual Image-Error Assessment Through Pairwise Preference (CVPR 2018) · 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
- Perceptual Texture Similarity Estimation: An Evaluation of Computational Features (TPAMI 2020 · similarity 0.90)
- ImageNet-trained CNNs are not biased towards texture: Revisiting feature reliance through controlled suppression (NeurIPS 2025 · similarity 0.87)
- Adaptive Image Quality Assessment via Teaching Large Multimodal Model to Compare (NeurIPS 2024 · similarity 0.87)
- Scaling and Masking: A New Paradigm of Data Sampling for Image and Video Quality Assessment (AAAI 2024 · similarity 0.86)
- Exploring Semantic Feature Discrimination for Perceptual Image Super-Resolution and Opinion-Unaware No-Reference Image Quality Assessment (CVPR 2025 · similarity 0.86)
- Toward Generalized Image Quality Assessment: Relaxing the Perfect Reference Quality Assumption (CVPR 2025 · 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.
