LoRA_Image_Preprocessor

LoRA Image Preprocessor

Overview

This Python tool serves as a dedicated preprocessor for images intended for LoRA (Low-Rank Adaptation) training. It automates the detection and blurring of faces or specific regions in both photos and videos. The inspiration for this tool stems from the Blurry Faces and remove-photo-background project, and I've tailored a modified version to cater to the unique requirements of LoRA training.

Demo

Credits

• Asmaa Mirkhan for the original Blurry Faces GitHub repository.
• aryadytm for the background removal code, sourced from remove-photo-background.