LatentCLR: A Contrastive Learning Approach for Unsupervised Discovery of Interpretable Directions
This addresses the need for unsupervised semantic direction discovery in GANs, enabling more practical and subtle image editing applications.
The paper tackled the problem of discovering interpretable directions in GAN latent spaces without manual annotations, proposing a contrastive learning approach that achieves performance comparable to state-of-the-art methods.
Recent research has shown that it is possible to find interpretable directions in the latent spaces of pre-trained Generative Adversarial Networks (GANs). These directions enable controllable image generation and support a wide range of semantic editing operations, such as zoom or rotation. The discovery of such directions is often done in a supervised or semi-supervised manner and requires manual annotations which limits their use in practice. In comparison, unsupervised discovery allows finding subtle directions that are difficult to detect a priori. In this work, we propose a contrastive learning-based approach to discover semantic directions in the latent space of pre-trained GANs in a self-supervised manner. Our approach finds semantically meaningful dimensions comparable with state-of-the-art methods.