Chao Chen

• 
  Topological Loss for Deep Generative Models
  Chao Chen
  Existing generative models (GAN or VAE) focus on generating realistic images based on CNN-derived image features, but fail to preserve the structural properties of real images. This can be fatal in applications where the underlying structure (e.g., neurons, vessels, membranes, and road networks) of the image carries crucial semantic meaning. We propose a novel GAN model that learns the topology of real images, i.e., connectedness and loopy-ness. In particular, we introduce a topological GAN loss that bridges the gap between synthetic image distribution and real image distribution in the topological feature space. By optimizing this loss, the generator produces images with the same structural topology as real images. We also propose new GAN evaluation metrics that measure the topological realism of the synthetic images. We show in experiments that our method generates synthetic images with realistic topology. We also highlight the increased performance that our method brings to downstream tasks such as segmentation.