This repository is unofficial implementation of RF Inversion, "Semantic Image Inversion and Editing using Stochastic Rectified Differential Equations". https://rf-inversion.github.io/
python main.py --image examples/man.jpg --prompt "A portrait of a man wearing glasses" --output outputs/man_glasses_eta09_s6e20.jpg --eta 0.9 --start_timestep 6 --stop_timestep 20
etaが大きいと元の画像に近くなる
start_timestepとstop_timestepはetaを有効にするtimestepで
0 <= start_timestep < stop_timestep < num_inference_step
コントロールが難しい