Updated version: option render factor added (only commandline version)
New models for use with render factor:
https://drive.google.com/drive/folders/1bU9Zj7zGVEujIzvDTb1b9cyWU3s__WQR?usp=sharing
.
Simple image and video colorization using onnx converted deoldify model.
Easy to install. Can be run on CPU or nVidia GPU
ffmpeg for video colorzation required.
Added floating point 16 model for 100% faster inference and simple GUI version.
For inference run:
Image: python image.py --source_image "image.jpg"
Video: python video.py --source "video.mp4" --result "video_colorized.mp4" --audio
Image example:
video_360.mp4
video_colorized_360.mp4
original deoldify: