Freeze the Discriminator: a Simple Baseline for Fine-Tuning GANs

25 Feb 2020  ·  Sangwoo Mo, Minsu Cho, Jinwoo Shin ·

Generative adversarial networks (GANs) have shown outstanding performance on a wide range of problems in computer vision, graphics, and machine learning, but often require numerous training data and heavy computational resources. To tackle this issue, several methods introduce a transfer learning technique in GAN training. They, however, are either prone to overfitting or limited to learning small distribution shifts. In this paper, we show that simple fine-tuning of GANs with frozen lower layers of the discriminator performs surprisingly well. This simple baseline, FreezeD, significantly outperforms previous techniques used in both unconditional and conditional GANs. We demonstrate the consistent effect using StyleGAN and SNGAN-projection architectures on several datasets of Animal Face, Anime Face, Oxford Flower, CUB-200-2011, and Caltech-256 datasets. The code and results are available at https://github.com/sangwoomo/FreezeD.

PDF Abstract

Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
10-shot image generation Babies FreezeD FID 96.25 # 5

Methods