Probing TryOnGAN
This work provides incremental insights into the training and conditioning mechanisms of virtual try-on systems for fashion applications.
The researchers reproduced the TryOnGAN virtual try-on system and analyzed its training dynamics, finding that transfer learning provides only temporary benefits and that pose conditioning via concatenation yields better performance than other methods.
TryOnGAN is a recent virtual try-on approach, which generates highly realistic images and outperforms most previous approaches. In this article, we reproduce the TryOnGAN implementation and probe it along diverse angles: impact of transfer learning, variants of conditioning image generation with poses and properties of latent space interpolation. Some of these facets have never been explored in literature earlier. We find that transfer helps training initially but gains are lost as models train longer and pose conditioning via concatenation performs better. The latent space self-disentangles the pose and the style features and enables style transfer across poses. Our code and models are available in open source.