Lecture 24 : SOTA Image PreProcessing Scripts For Stable Diffusion Training – Auto Subject Crop & Face Focus
One of the most important aspect of Stable Diffusion training is the preparation of training images. In this tutorial video I will show you how to fully automatically preprocess training images with perfect zoom, crop and resize. These scripts will hugely improve your training success and accuracy.
Scripts Download Link ⤵️
Massive and excellent quality classification / regularization preprocessed and raw images datasets for both man and woman ⤵️
The scripts are utilizing Yolo V7 and Retina Face AI models. These are the very best state of the art AI models. Moreover they are very well coded. They will work super optimized and accurate.
With these scripts with 1 click you will preprocess all of your training images into the desired target resolution with very possible way and quality.
SECourses Discord To Get Full Support ⤵️
My LinkedIn ⤵️
My Twitter ⤵️
My Instagram ⤵️
My Medium ⤵️
My CivitAI ⤵️
0:00 Introduction to the very best training images preprocessing scripts for Stable Diffusion training
5:48 How to install and use auto cropper / zoom-in script – first stage
7:25 Which file to edit and set processing folder path for cropper script
7:52 How to add or remove new aspect ratios
10:18 Full cropping supported classes of Yolo V7
11:42 How to install and run image resizer / downscaler with face focus with Retina Face
12:51 How to edit and run resizer script for perfect downscale and preparation of training images
14:37 Comparison of raw images vs final processed images