A project that can compose an image into a cartoon or the style of another image by using pre-trained deep learning models.
A GLSurfaceView is used for rendering image into view and the image data is transfered from view to ML models, and vice versa, through a pixel buffer rather than bitmap. This GLSurfaceView provides a OpenGL ES context which will be effective in further pre/post processing.
- Image resources as Styled Images or Sample Input are picked from Unsplash
- Thank you to @MasayukiSuda for his work on Mp4Composer-android project where OpenGL ES components are organized very well.
- Thanks to Google TensorFlow team and the people who review and publish TFLite version of ML models.