This project showcases an extension of RIFE for applications in animation. For a this project submission and talk of the underlying theory, read FrameInterpforAnimation.pdf.

To see a log I've written of development, check out corlene_notes.md (tbh it's a better README than my README). This will also include my detailed description of packages needed for install and bugs I ran into, though theoretically requirements.txt should work on its own.

If you'd like to read RIFE's original readme, check out RIFE-README.md. This repo is forked from https://github.com/hzwer/ECCV2022-RIFE

Data used in this project was collected by the people behind AnimeInterp. Their github can be found at https://github.com/lisiyao21/AnimeInterp

If you'd like to see the trained models, check out the final_models folder. They will be ready to run using inference_img.py

Alternatively, if you'd like to generate the example images in the paper, try running the shell script get_all_model_outputs.sh

Thanks for checking this out :>