Research map / Image generation / GANs
Generative Adversarial Minority Oversampling
Sankha Subhra Mullick, Shounak Datta, Swagatam Das
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 a discriminator D to the convex generator (GAMO vs GAMO\D) (ablation) | GAMO · MNIST (IR 100, LS loss) | ACSA: 0.88 -> 0.91 | Improved |
| add a discriminator D to the convex generator (GAMO vs GAMO\D) (ablation) | GAMO · Fashion-MNIST (IR 100, LS loss) | ACSA: 0.82 -> 0.83 | No clear effect |
| add a discriminator D to the convex generator (GAMO vs GAMO\D) (ablation) | GAMO · CIFAR10 (IR 56.25) | GM: 0.40 -> 0.43 | Improved |
| use convex generator instead of conditional generator cG (both with D) (ablation) | GAMO · MNIST (IR 100, LS loss) | ACSA: 0.85 -> 0.91 | Improved |
| add discriminator D to conditional generator cG (ablation) | cG+CN · MNIST (IR 100, LS loss) | GM: 0.85 -> 0.82 | Hurt |
| train generator adversarially against classifier (cG+CN) instead of cGAN+CN (ablation) | cGAN+CN · MNIST (IR 100, LS loss) | ACSA: 0.89 -> 0.86 | Hurt |
| use least-square loss instead of cross entropy (ablation) | GAMO · MNIST | ACSA: 0.89 -> 0.91 | Improved |
| use least-square loss instead of cross entropy (ablation) | GAMO · Fashion-MNIST | ACSA: 0.82 -> 0.83 | Improved |
| use GAMO instead of baseline classifier | CN · SUN397 (50 classes) | ACSA: 0.26 -> 0.32 | Improved |
| use GAMO instead of data augmentation | Augment+CN · SVHN (IR 56.25) | ACSA: 0.69 -> 0.76 | Improved |
| use GAMO2pix instead of cDCGAN | image generation · Fashion-MNIST | FID: 5.57 -> 0.75 | Improved |
| use GAMO instead of baseline classifier at larger data size | CN · CelebA-Large (test) | ACSA: 0.71 -> 0.75 | Improved |
Closest papers by meaning
- Synthetic Tabular Data Generation for Imbalanced Classification: The Surprising Effectiveness of an Overlap Class (AAAI 2025 · similarity 0.87)
- Balance, Imbalance, and Rebalance: Understanding Robust Overfitting from a Minimax Game Perspective (NeurIPS 2023 · similarity 0.86)
- Multi-objective training of Generative Adversarial Networks with multiple discriminators (ICML 2019 · similarity 0.85)
- Deep Generative Model for Robust Imbalance Classification (CVPR 2020 · similarity 0.85)
- Concentration and excess risk bounds for imbalanced classification with synthetic oversampling (NeurIPS 2025 · similarity 0.85)
- Restoring balance: principled under/oversampling of data for optimal classification (ICML 2024 · similarity 0.84)
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.
