Research map / Image generation / Image restoration
Image Aesthetic Assessment Based on Pairwise Comparison A Unified Approach to Score Regression, Binary Classification, and Personalization
Jun-Tae Lee, Chang-Su Kim
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 ternary instead of 5-ary quantization of score ratio (ablation) | aesthetic comparator · AVA score regression | Spearman rho: 0.791 -> 0.918 | Improved |
| use ternary instead of 7-ary quantization of score ratio (ablation) | aesthetic comparator · AVA score regression | Spearman rho: 0.779 -> 0.918 | Improved |
| use ternary instead of 5-ary quantization (ablation) | aesthetic comparator · AADB score regression | MASD: 0.1713 -> 0.1141 | Improved |
| use ternary instead of 7-ary quantization (ablation) | aesthetic comparator · AADB score regression | Spearman rho: 0.867 -> 0.879 | Improved |
| add pairwise comparison with reference images to baseline network | baseline network · AVA binary classification | accuracy (%): 78.7 -> 91.5 | Improved |
| use pairwise-comparison classifier instead of A-Lamp | proposed · AVA binary classification | accuracy (%): 82.5 -> 91.5 | Improved |
| use proposed regressor instead of Reg-Net | proposed · AVA score regression | MASD: 0.0582 -> 0.0229 | Improved |
| use proposed regressor instead of PAC-Net | proposed · AVA score regression | Spearman rho: 0.871 -> 0.918 | Improved |
| add 10 personal reference images (ablation) | proposed · FLICKER-AES | Spearman rho: 0.668 -> +0.040 | Improved |
| add 100 personal reference images instead of 10 (ablation) | proposed · FLICKER-AES | Spearman rho: +0.040 -> +0.044 | Mixed |
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.
- A-Lamp — A-Lamp: Adaptive Layout-Aware Multi-Patch Deep Convolutional Neural Network for Photo Aesthetic Assessment (CVPR 2017) · better on 1 of 1 shared result
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
- Scaling Up Personalized Image Aesthetic Assessment via Task Vector Customization (ECCV 2024 · similarity 0.90)
- Revisiting Image Aesthetic Assessment via Self-Supervised Feature Learning (AAAI 2020 · similarity 0.88)
- Advancing Comprehensive Aesthetic Insight with Multi-Scale Text-Guided Self-Supervised Learning (AAAI 2025 · similarity 0.87)
- Bridging Cognitive Gap: Hierarchical Description Learning for Artistic Image Aesthetics Assessment (AAAI 2026 · similarity 0.86)
- ArtiMuse: Fine-Grained Image Aesthetics Assessment with Joint Scoring and Expert-Level Understanding (CVPR 2026 · similarity 0.85)
- Unlocking the Essence of Beauty: Advanced Aesthetic Reasoning with Relative-Absolute Policy Optimization (ICLR 2026 · similarity 0.85)
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.
