The Frame Interpolation project introduces an approach to enhance video playback smoothness by intelligently increasing the frame rate. Utilizing a convolutional neural network techniques built with PyTorch, this tool interpolates and inserts new frames between existing ones, effectively transforming choppy videos into fluid motion sequences.
Make sure to have Python (version 3.6 or higher) and PyTorch installed. Then, run the following command to install the required dependencies:
pip install -r requirements.txt
halo.mp4
halo_48fps.mp4
Execute runner.py to begin. Select between preprocessing data or generating an interpolated video. To train the model, preprocess your data first, followed by running model_training.py. To generate an interpolated video, make sure to have the input video in the videos folder and follow instructions.