Cartoonize your world!

Due credits to the incredible paper - Learning to Cartoonize Using White-box Cartoon Representations

Official implementation of the paper by the author - Github Link

• We are currently processing only upto 10 second videos, if you happened to upload a video greater than 10 seconds, only first 10 seconds will be considered.
• Video File Size Limitation: 30MB
• Image File Formats Supported: jpeg, png
• Video File Formats Supported: mp4, webm, avi, mkv
• GIF/TIFF Images are not supported.

Some of the areas where we think it could be applied to -

• Churn out quick prototypes or sprites for animes, cartoons and games
• Since it subdues facial features and information in general, it can be used to generate minimal art
• Games can import short cut scenes very easily without using motion-capture
• Can be modelled as an assistant to graphic designers or animators.

Cake and food assortment photos are OC (Original Content). Other than that -

• Emma Watson Image: Independent UK
• Rick Astley - Never Gonna Give You Up (Original Video): YouTube Link
• Avengers (Original Video): YouTube Link
• Joey (Original Video): YouTube Link

The author of the above mentioned paper can probably indulge you in some detailed resoursces. Other than that here are some we found -

• CartoonGAN
  • Paper
  • Github Link
• AnimeGAN
  • Paper
  • Github Link

Honestly we thought this was a cool application of GAN but didn't find any demo available. Our friends wanted to try it so we made a quick demo for images; which then later got extended to videos.

Also we wanted to learn the deployment architecture which would allow us to serve such power hungry inference in the least money hogging method possible. (Blog post coming soon!)

This is a Flask app which resides on a Cloud Run instance with Cloud Storage integration. We've leveraged Algorithmia's community cloud AI layer for our inference.

Yes it is! - Github Link