Inspiration
What would it feel like watching live action adaptation of your favourite anime shows and movies?
What it does
Translates Anime scenes to Photorealistic Scenes using Generative Deep Learning models.
How we built it
Inspired from 2022 paper "ITTR: Unpaired Image-to-Image Translation with Transformers" , used a pretrained ViT as backbone for encoder.
Challenges we ran into
Data acquisition
Accomplishments that we're proud of
Close to photorealism. Learning along the way !
What we learned
Successfully built and trained a sate of the art vision-transformer model.
What's next for Anime-IRL
Even more advanced Deep Learning based generative models. Pushing the state of the art !
Log in or sign up for Devpost to join the conversation.