Research map / Image generation / GANs
Introspective Neural Networks for Generative Modeling
Justin Lazarow, Long Jin, Zhuowen Tu
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 INNg instead of DCGAN features for unsupervised feature learning | DCGAN · SVHN semi-supervised (L2-SVM on 1000 labeled examples) | test error rate: 33.13% -> 32.81% | Improved |
| use a cascade of many classifiers (INNg) instead of a single classifier (INNg-single) (ablation) | INNg-single · texture, CelebA and SVHN generation (qualitative) | generative quality | Improved |
| use INNg (60 classifiers) instead of Gatys et al. and DCGAN for texture synthesis | Gatys et al. · texture synthesis (qualitative) | visual quality | Improved |
| use INNg backprop synthesis instead of GAN single forward pass | GAN · image synthesis | speed and model size: about 10 seconds per 50 images of 64x64; 10-60 classifiers stored | Hurt |
Closest papers by meaning
- Introspective Classification with Convolutional Nets (NeurIPS 2017 · similarity 0.90)
- Learning FRAME Models Using CNN Filters (AAAI 2016 · similarity 0.87)
- Counterfactual Generative Networks (ICLR 2021 · similarity 0.87)
- Intriguing Properties of Generative Classifiers (ICLR 2024 · similarity 0.86)
- Direct Discriminative Optimization: Your Likelihood-Based Visual Generative Model is Secretly a GAN Discriminator (ICML 2025 · similarity 0.85)
- Machine Unlearning for Image-to-Image Generative Models (ICLR 2024 · 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.
