StyleMC: Multi-Channel Based Fast Text-Guided Image Generation and Manipulation
This addresses the need for efficient text-driven image editing for users in creative and AI applications, though it is incremental as it builds on existing CLIP and StyleGAN2 frameworks.
The paper tackles the problem of slow text-guided image manipulation in GANs by introducing StyleMC, a method that reduces training time to a few seconds per prompt while maintaining attribute control, achieving competitive results with state-of-the-art methods.
Discovering meaningful directions in the latent space of GANs to manipulate semantic attributes typically requires large amounts of labeled data. Recent work aims to overcome this limitation by leveraging the power of Contrastive Language-Image Pre-training (CLIP), a joint text-image model. While promising, these methods require several hours of preprocessing or training to achieve the desired manipulations. In this paper, we present StyleMC, a fast and efficient method for text-driven image generation and manipulation. StyleMC uses a CLIP-based loss and an identity loss to manipulate images via a single text prompt without significantly affecting other attributes. Unlike prior work, StyleMC requires only a few seconds of training per text prompt to find stable global directions, does not require prompt engineering and can be used with any pre-trained StyleGAN2 model. We demonstrate the effectiveness of our method and compare it to state-of-the-art methods. Our code can be found at http://catlab-team.github.io/stylemc.