Adversarial Semantic Augmentation for Training Generative Adversarial Networks under Limited Data

TL;DR

This paper introduces an adversarial semantic augmentation method to improve GAN training with limited data by enlarging training sets at the semantic level, leading to better image synthesis quality.

Contribution

The paper proposes a novel semantic augmentation technique that estimates semantic feature transformations via covariance matrices, avoiding data distribution leakage and reducing computational overhead.

Findings

Improves GAN synthesis quality in low-data regimes

Enhances performance on both few-shot and large-scale datasets

Maintains computational efficiency during training

Abstract

Generative adversarial networks (GANs) have made remarkable achievements in synthesizing images in recent years. Typically, training GANs requires massive data, and the performance of GANs deteriorates significantly when training data is limited. To improve the synthesis performance of GANs in low-data regimes, existing approaches use various data augmentation techniques to enlarge the training sets. However, it is identified that these augmentation techniques may leak or even alter the data distribution. To remedy this, we propose an adversarial semantic augmentation (ASA) technique to enlarge the training data at the semantic level instead of the image level. Concretely, considering semantic features usually encode informative information of images, we estimate the covariance matrices of semantic features for both real and generated images to find meaningful transformation directions. Such directions translate original features to another semantic representation, e.g., changing the backgrounds or expressions of the human face dataset. Moreover, we derive an upper bound of the expected adversarial loss. By optimizing the upper bound, our semantic augmentation is implicitly achieved. Such design avoids redundant sampling of the augmented features and introduces negligible computation overhead, making our approach computation efficient. Extensive experiments on both few-shot and large-scale datasets demonstrate that our method consistently improve the synthesis quality under various data regimes, and further visualized and analytic results suggesting satisfactory versatility of our proposed method.