Research map / Label-efficient and robust learning / Recognition and long-tail
The Shape of Art History in the Eyes of the Machine
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 ImageNet pre-training and fine-tune on WikiArt instead of training from scratch | AlexNet · WikiArt | classification accuracy: 48.1% -> 58.2% | Helped |
| Use ImageNet pre-training and fine-tune on WikiArt instead of training from scratch | AlexNet+2 · WikiArt | classification accuracy: 47.2% -> 58.3% | Helped |
| Use ImageNet pre-training and fine-tune on WikiArt instead of training from scratch | VGGNet · WikiArt | classification accuracy: 51.6 % -> 60.1% | Helped |
| Use ImageNet pre-training and fine-tune on WikiArt instead of training from scratch | VGGNet+2 · WikiArt | classification accuracy: 55.2% -> 62.8% | Helped |
| Add two reduced fully connected layers (ablation) | AlexNet · WikiArt | classification accuracy: 58.2% -> 58.3% | No clear effect |
| Add two reduced fully connected layers (ablation) | VGGNet · WikiArt | classification accuracy: 60.1% -> 62.8% | Helped |
| Add two reduced fully connected layers (ablation) | ResNet-152 · WikiArt | classification accuracy: 63.7% -> 60.2% | Hurt |
| Add two reduced fully connected layers (ablation) | AlexNet trained from scratch · WikiArt | classification accuracy: 48.1% -> 47.2% | Hurt |
| Add two reduced fully connected layers (ablation) | VGGNet trained from scratch · WikiArt | classification accuracy: 51.6 % -> 55.2% | Helped |
| Add two reduced fully connected layers (ablation) | ResNet-50 trained from scratch · WikiArt | classification accuracy: 45.0% -> 48.0% | Helped |
| Increase network depth from AlexNet to ResNet-152 (ablation) | Style classification models · WikiArt | classification accuracy: 48.1% -> 63.7%; ResNet-50 performed worse than AlexNet | Mixed |
| Add two reduced fully connected layers (ablation) | Fine-tuned AlexNet · Artchive visualization set | PCA subspace dimension retaining 95% variance: 201 -> 9 | Helped |
Closest papers by meaning
- Gender Artifacts from Art History to Text-to-Image Generation (NeurIPS 2026 · similarity 0.86)
- APDDv2: Aesthetics of Paintings and Drawings Dataset with Artist Labeled Scores and Comments (NeurIPS 2024 · similarity 0.83)
- Stylebreeder: Exploring and Democratizing Artistic Styles through Text-to-Image Models (NeurIPS 2024 · similarity 0.83)
- FiVA: Fine-grained Visual Attribute Dataset for Text-to-Image Diffusion Models (NeurIPS 2024 · similarity 0.82)
- Bridging Cognitive Gap: Hierarchical Description Learning for Artistic Image Aesthetics Assessment (AAAI 2026 · similarity 0.82)
- ArtBank: Artistic Style Transfer with Pre-trained Diffusion Model and Implicit Style Prompt Bank (AAAI 2024 · similarity 0.82)
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.