AI DAIN The final Agni Kai

I had come across various interesting AI's, in a former project before quarantine I was working with noise cancelling. I was still very much interested in other option and decided to just wanted to try things out and see what happens in my spare time.

with this experiment I used the AI DAIN to process a video, This scrips tries to add extra frames by using AI which is constantly learning. It makes of each frame a PNG file and then tries to generate interpolated frames. This together with the original frames are editten together to create this 50fps version of the original video.

I used Avatar The Last Airbender a Final Agni Kai for this projects since I found this scene as a child and still now very nice. I had recently come across it again and had a lovely time. This is also an animation which is a little bit lighter for the AI. Even with a very strong computer this takes a lot of time and power.


We propose the Depth-Aware video frame INterpolation (DAIN) model to explicitly detect the occlusion by exploring the depth cue. We develop a depth-aware flow projection layer to synthesize intermediate flows that preferably sample closer objects than farther ones. Video frame interpolation aims to synthesize non-existent frames in-between the original frames.

Depth-Aware Video Frame Interpolation (CVPR 2019). Contribute to baowenbo/DAIN development by creating an account on GitHub.

Sadly due to CopyRight the video's were monitized on Youtube so I have posted the on my own website instead.