GANzilla: User-Driven Direction Discovery in Generative Adversarial Networks
This addresses the challenge for non-expert users in controlling GAN-generated data, offering a complementary approach to algorithm-driven methods, though it is incremental as it builds on existing scatter/gather techniques.
The authors tackled the problem of non-expert users struggling to control GAN outputs due to its 'black box' nature by developing GANzilla, a user-driven tool that enables iterative discovery of editing directions using scatter/gather techniques. In a study with 12 participants, users successfully discovered directions for both closed-ended tasks (matching examples) and open-ended tasks (e.g., making faces happier) with diversity across individuals.
Generative Adversarial Network (GAN) is widely adopted in numerous application areas, such as data preprocessing, image editing, and creativity support. However, GAN's 'black box' nature prevents non-expert users from controlling what data a model generates, spawning a plethora of prior work that focused on algorithm-driven approaches to extract editing directions to control GAN. Complementarily, we propose a GANzilla: a user-driven tool that empowers a user with the classic scatter/gather technique to iteratively discover directions to meet their editing goals. In a study with 12 participants, GANzilla users were able to discover directions that (i) edited images to match provided examples (closed-ended tasks) and that (ii) met a high-level goal, e.g., making the face happier, while showing diversity across individuals (open-ended tasks).