FLAIR

This is the official Repo for WACV2025 paper "FLAIR: A Conditional Diffusion Framework with Applications to Face Video Restoration" (paper)

Installation

pip install -r requirements.txt
pip install flash-attn --no-build-isolation

Download Pre-trained Models

Google Drive

Place them in ./checkpoints

Usage

python scripts/video_sample.py gaussian-demo --device cuda

python scripts/video_sample.py x8-bicubic-demo --device cuda

python scripts/video_sample.py x16-bicubic-demo --device cuda

python scripts/video_sample.py jpeg-demo --device cuda