Research map / Label-efficient and robust learning / Recognition and long-tail
Simple Disentanglement of Style and Content in Visual Representations
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 |
|---|---|---|---|
| Post-process ResNet-50 features with PISCO (λ=1) | ResNet-50 · ImageNet dog sketch | top-1 accuracy: 0.516 -> 0.546 | Helped |
| Post-process ResNet-50 features with PISCO (λ=1) | ResNet-50 · ImageNet woman sketch | top-1 accuracy: 0.478 -> 0.518 | Helped |
| Post-process ResNet-50 features with PISCO (λ=1) | ResNet-50 · ImageNet Picasso dog | top-1 accuracy: 0.445 -> 0.500 | Helped |
| Post-process ResNet-50 features with PISCO (λ=1) | ResNet-50 · ImageNet Picasso self-portrait | top-1 accuracy: 0.474 -> 0.514 | Helped |
| Post-process ResNet-50 features with PISCO (λ=1) | ResNet-50 · ImageNet none | top-1 accuracy: 0.757 -> 0.749 | Hurt |
| Post-process MAE-ViT-Base features with PISCO (λ=1) | MAE-ViT-Base · ImageNet dog sketch | top-1 accuracy: 0.530 -> 0.575 | Helped |
| Post-process MAE-ViT-Base features with PISCO (λ=1) | MAE-ViT-Base · ImageNet Picasso dog | top-1 accuracy: 0.472 -> 0.519 | Helped |
| Post-process MAE-ViT-Base features with PISCO (λ=1) | MAE-ViT-Base · ImageNet Picasso self-portrait | top-1 accuracy: 0.512 -> 0.558 | Helped |
| Post-process MAE-ViT-Base features with PISCO (λ=1) | MAE-ViT-Base · ImageNet woman sketch | top-1 accuracy: 0.504 -> 0.550 | Helped |
| Post-process MAE-ViT-Base features with PISCO (λ=1) | MAE-ViT-Base · ImageNet none | top-1 accuracy: 0.811 -> 0.818 | Helped |
Closest papers by meaning
- Unsupervised Disentanglement of Content and Style via Variance-Invariance Constraints (ICLR 2025 · similarity 0.90)
- Factorized Diffusion Autoencoder for Unsupervised Disentangled Representation Learning (AAAI 2024 · similarity 0.87)
- Disentangled Representation Learning via Modular Compositional Bias (NeurIPS 2025 · similarity 0.87)
- Transferring disentangled representations: bridging the gap between synthetic and real images (NeurIPS 2024 · similarity 0.87)
- Disentanglement Beyond Static vs. Dynamic: A Benchmark and Evaluation Framework for Multi-Factor Sequential Representations (NeurIPS 2025 · similarity 0.87)
- Can Diffusion Models Disentangle? A Theoretical Perspective (NeurIPS 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.