Research map / Image generation / GANs
AE-OT-GAN: Training GANs from data specific latent distribution
Dongsheng An, Yang Guo, Min Zhang, Xin Qi, Na Lei, Xianfang Gu
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 GAN training with content/feature losses on top of AE-OT decoder | AE-OT · CelebA (ResNet) | FID: 28.6 -> 7.8 | Improved |
| add GAN training on top of AE-OT decoder | AE-OT · CelebA (standard) | FID: 24.3 -> 11.2 | Improved |
| add GAN training on top of AE-OT decoder | AE-OT · CIFAR10 (ResNet) | FID: 28.5 -> 17.1 | Improved |
| add GAN training on top of AE-OT decoder | AE-OT · CIFAR10 (standard) | FID: 34.2 -> 25.2 | Improved |
| sample GAN latents from the AE-OT latent distribution instead of Gaussian (vs WGAN-GP) | GAN · CIFAR10 (standard) | FID: 40.2 -> 25.2 | Improved |
| use AE-OT-GAN instead of WGAN-QC | GAN · CelebA (ResNet) | FID: 12.9 -> 7.8 | Improved |
| use AE-OT-GAN instead of SNGAN | GAN · CIFAR10 (standard) | FID: 25.5 -> 25.2 | No clear effect |
| use AE-OT-GAN instead of WGAN-div | GAN · CIFAR10 (ResNet) | FID: 18.1 -> 17.1 | Improved |
| use AE-OT-GAN instead of WGAN-QC | GAN · CelebA-HQ | FID: 7.7 -> 7.2 | Improved |
| use AE-OT-GAN instead of earlier best MNIST result | GAN with InfoGAN architecture · MNIST | FID: 6.4 -> 3.2 | Improved |
Later papers that compared with this method
1 later accepted paper reports AE-OT-GAN as a baseline in their results tables; 0 of them report a better number on most of the results they share with it.
- Generative Modeling with Optimal Transport Maps (ICLR 2022 · better on 1 of 2 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- AE-OT: A NEW GENERATIVE MODEL BASED ON EXTENDED SEMI-DISCRETE OPTIMAL TRANSPORT (ICLR 2020 · similarity 0.92)
- Improving Generative Adversarial Networks via Adversarial Learning in Latent Space (NeurIPS 2022 · similarity 0.88)
- Complexity Matters: Rethinking the Latent Space for Generative Modeling (NeurIPS 2023 · similarity 0.87)
- Adversarial Latent Autoencoders (CVPR 2020 · similarity 0.87)
- CLR-GAN: Improving GANs Stability and Quality via Consistent Latent Representation and Reconstruction (ECCV 2024 · similarity 0.87)
- MCGAN: Enhancing GAN Training with Regression-Based Generator Loss (AAAI 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.
