Semi-Supervised StyleGAN for Disentanglement Learning

Disentanglement learning is crucial for obtaining disentangled representations and controllable generation. Current disentanglement methods face several inherent limitations: difficulty with high-resolution images, primarily focusing on learning disentangled representations, and nonidentifiability due to the unsupervised setting. To alleviate these limitations, we design new architectures and loss functions based on StyleGAN (Karras et al., 2019), for semi-supervised highresolution disentanglement learning. We create two complex high-resolution synthetic datasets for systematic testing. We investigate the impact of limited supervision and find that using only 0.25%∼2.5% of labeled data is sufficient for good disentanglement on both synthetic and real datasets. We propose new metrics to quantify generator controllability, and observe there may exist a crucial trade-off between disentangled representation learning and controllable generation. We also consider semantic fine-grained image editing to achieve better generalization to unseen images.