Timezone: »

 
Poster
Data-Efficient Instance Generation from Instance Discrimination
Ceyuan Yang · Yujun Shen · Yinghao Xu · Bolei Zhou

Wed Dec 08 12:30 AM -- 02:00 AM (PST) @
in Poster Session 3 »
Generative Adversarial Networks (GANs) have significantly advanced image synthesis, however, the synthesis quality drops significantly given a limited amount of training data. To improve the data efficiency of GAN training, prior work typically employs data augmentation to mitigate the overfitting of the discriminator yet still learn the discriminator with a bi-classification ($\textit{i.e.}$, real $\textit{vs.}$ fake) task. In this work, we propose a data-efficient Instance Generation ($\textit{InsGen}$) method based on instance discrimination. Concretely, besides differentiating the real domain from the fake domain, the discriminator is required to distinguish every individual image, no matter it comes from the training set or from the generator. In this way, the discriminator can benefit from the infinite synthesized samples for training, alleviating the overfitting problem caused by insufficient training data. A noise perturbation strategy is further introduced to improve its discriminative power. Meanwhile, the learned instance discrimination capability from the discriminator is in turn exploited to encourage the generator for diverse generation. Extensive experiments demonstrate the effectiveness of our method on a variety of datasets and training settings. Noticeably, on the setting of $2K$ training images from the FFHQ dataset, we outperform the state-of-the-art approach with 23.5\% FID improvement.
Keywords: Machine Learning · Generative Model

Author Information

Ceyuan Yang (The Chinese University of Hong Kong)
Yujun Shen (ByteDance Inc.)
Yinghao Xu (Chinese University of Hong Kong)
Bolei Zhou (Massachusetts Institute of Technology)

More from the Same Authors