AniimeGAN v2 Sample

  • Description
    Generative neural networks unveil a new era of art. With such models artists can enhance their work by improving its quality, adding unique details and styles. AnimeGANv2 is open source model styling photos or videos in a selected anime style. We used it to demonstrate how Savant can play in another role, not only to detect, classify or segment data but also to generate a new video based on the original.
  • Key Features
    —AnimeGANv2 generative model
  • End-To-End Performance
    — 4.5 FPS on Nvidia A4000