SwipeGANSpace: Swipe-to-Compare Image Generation via Efficient Latent Space Exploration

TL;DR

SwipeGANSpace introduces an efficient method for generating user-preferred images by exploring StyleGAN's latent space through simple swipe interactions, leveraging PCA and multi-armed bandits to adapt to dynamic user preferences.

Contribution

The paper presents a novel interactive approach combining PCA and multi-armed bandits for efficient latent space exploration based on user swipes, enabling dynamic preference modeling.

Findings

More efficient than baseline methods in generating preferred images.

Captures and adapts to changing user preferences during interaction.

Reveals the dynamic nature of user preferences in image generation.

Abstract

Generating preferred images using generative adversarial networks (GANs) is challenging owing to the high-dimensional nature of latent space. In this study, we propose a novel approach that uses simple user-swipe interactions to generate preferred images for users. To effectively explore the latent space with only swipe interactions, we apply principal component analysis to the latent space of the StyleGAN, creating meaningful subspaces. We use a multi-armed bandit algorithm to decide the dimensions to explore, focusing on the preferences of the user. Experiments show that our method is more efficient in generating preferred images than the baseline methods. Furthermore, changes in preferred images during image generation or the display of entirely different image styles were observed to provide new inspirations, subsequently altering user preferences. This highlights the dynamic nature of user preferences, which our proposed approach recognizes and enhances.