Research map / Image generation / Autoregressive and latent generation
Predicting Aesthetic Score Distribution Through Cumulative Jensen-Shannon Divergence
Xin Jin, Le Wu, Xiaodong Li, Siyu Chen, Siwei Peng, Jingying Chi, Shiming Ge, Chenggen Song, Geng Zhao
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 kurtosis-based reliability weighting to CJS loss (ablation) | CJS-CNN · AVA test | CJS divergence: 0.041 -> 0.040 | Improved |
| Replace kurtosis reliability with rating-count reliability (ablation) | CJS-CNN · AVA test | CJS divergence: 0.040 -> 0.040 | No clear effect |
| Mix kurtosis and rating-count reliability with $\lambda = 0.1$ (ablation) | CJS-CNN · AVA test | CJS divergence: 0.040 -> 0.040 | No clear effect |
| Mix kurtosis and rating-count reliability with $\lambda = 0.3$ (ablation) | CJS-CNN · AVA test | CJS divergence: 0.040 -> 0.040 | No clear effect |
| Mix kurtosis and rating-count reliability with $\lambda = 0.5$ (ablation) | CJS-CNN · AVA test | CJS divergence: 0.040 -> 0.040 | No clear effect |
| Mix kurtosis and rating-count reliability with $\lambda = 0.7$ (ablation) | CJS-CNN · AVA test | CJS divergence: 0.040 -> 0.041 | Hurt |
| Mix kurtosis and rating-count reliability with $\lambda = 0.9$ (ablation) | CJS-CNN · AVA test | CJS divergence: 0.040 -> 0.040 | No clear effect |
Closest papers by meaning
- Image Aesthetic Assessment Assisted by Attributes through Adversarial Learning (AAAI 2019 · similarity 0.91)
- Effective Aesthetics Prediction With Multi-Level Spatially Pooled Features (CVPR 2019 · similarity 0.90)
- Photo Aesthetics Ranking Network with Attributes and Content Adaptation (ECCV 2016 · similarity 0.90)
- Bridging Cognitive Gap: Hierarchical Description Learning for Artistic Image Aesthetics Assessment (AAAI 2026 · similarity 0.85)
- Advancing Comprehensive Aesthetic Insight with Multi-Scale Text-Guided Self-Supervised Learning (AAAI 2025 · similarity 0.84)
- VADB: A Large-Scale Video Aesthetic Database with Professional and Multi-Dimensional Annotations (NeurIPS 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.
