Research map / Image generation / GANs
Least Squares Generative Adversarial Networks
Xudong Mao, Qing Li, Haoran Xie, Raymond Y.K. Lau, Zhen Wang, Stephen Paul Smolley
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 |
|---|---|---|---|
| replace sigmoid cross entropy loss with least squares loss | DCGAN without BN in G, Adam · LSUN bedroom | successful training runs: 5 of 10 succeed vs never successful | Improved |
| replace sigmoid cross entropy loss with least squares loss | DCGAN without BN in G and D, RMSProp · LSUN bedroom | image quality / mode collapse | Improved |
| replace sigmoid cross entropy loss with least squares loss | DCGAN without BN in G, RMSProp · LSUN bedroom | image quality | No clear effect |
| replace sigmoid cross entropy loss with least squares loss | DCGAN without BN in G and D, Adam · LSUN bedroom | mode collapse | No clear effect |
| replace sigmoid cross entropy loss with least squares loss | 3-layer FC GAN · 2D mixture of 8 Gaussians | mode coverage: regular GAN collapses from step 15k; LSGAN learns the mixture | Improved |
| use RMSProp instead of Adam (ablation) | regular GAN without BN in G · LSUN bedroom | successful learning | Improved |
| set a-b-c coding to b-c=1,b-a=2 versus 0-1 coding (c=b) (ablation) | LSGAN · LSUN | image quality | No clear effect |
Closest papers by meaning
- On the Effectiveness of Least Squares Generative Adversarial Networks (TPAMI 2018 · similarity 0.98)
- TRAINING GENERATIVE ADVERSARIAL NETWORKS VIA PRIMAL-DUAL SUBGRADIENT METHODS: A LAGRANGIAN PERSPECTIVE ON GAN (ICLR 2018 · similarity 0.89)
- Loss-Sensitive Generative Adversarial Networks on Lipschitz Densities (IJCV 2019 · similarity 0.89)
- Regularizing Generative Adversarial Networks Under Limited Data (CVPR 2021 · similarity 0.88)
- SGAN: An Alternative Training of Generative Adversarial Networks (CVPR 2018 · similarity 0.88)
- MCGAN: Enhancing GAN Training with Regression-Based Generator Loss (AAAI 2025 · similarity 0.87)
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.
