Research map / Image generation / GANs
MCL-GAN: Generative Adversarial Networks with Multiple Specialized Discriminators
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 |
|---|---|---|---|
| add MCL multiple specialized discriminators (M=10) | DCGAN · CIFAR-10 | FID: 37.70 -> 26.87 | Helped |
| add MCL multiple specialized discriminators (M=10) | StyleGAN2 · CIFAR-10 | FID: 9.06 -> 7.13 | Helped |
| use MCL-GAN instead of GMAN (M=5 vs 5) | DCGAN · CelebA | FID: 31.66 -> 16.88 | Helped |
| use MCL-GAN instead of GMAN | DCGAN · Fashion-MNIST | precision: 0.955 -> 0.929 | Mixed |
| assign real samples to expert discriminators by MCL rather than minimum-score selection (ablation) | MCL-GAN · MNIST | recall: 0.913 -> 0.985 | Helped |
| assign real samples by MCL rather than random selection (ablation) | MCL-GAN · CelebA | recall: 0.930 -> 0.955 | Helped |
| replace ground-truth class-label assignment with MCL assignment (ablation) | MCL-GAN · MNIST | recall: 0.978 -> 0.985 | No clear effect |
| add MCL to DRIT | DRIT · Dog->Cat | FID: 62.85 -> 20.61 | Helped |
| add MCL to DRIT | DRIT · Cat->Dog | LPIPS: 0.245 -> 0.511 | Helped |
| add MCL to DRIT | DRIT · Summer->Winter | NDB: 25.60 -> 25.40 | No clear effect |
| add MCL to StackGAN++ | StackGAN++ · CUB-200-2011 | LPIPS: 0.36 -> 0.63 | Helped |
| add MCL on top of mode-seeking regularizer in StackGAN++ | StackGAN++ + MS · CUB-200-2011 | FID: 25.53 -> 25.44 | No clear effect |
Closest papers by meaning
- Dual Discriminator Generative Adversarial Nets (NeurIPS 2017 · similarity 0.92)
- MGAN: Training Generative Adversarial Nets with Multiple Generators (ICLR 2018 · similarity 0.92)
- Triple Generative Adversarial Nets (NeurIPS 2017 · similarity 0.91)
- Generative Multi-Adversarial Networks (ICLR 2017 · similarity 0.91)
- DO-GAN: A Double Oracle Framework for Generative Adversarial Networks (CVPR 2022 · similarity 0.91)
- MCGAN: Enhancing GAN Training with Regression-Based Generator Loss (AAAI 2025 · similarity 0.89)
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.