In this video, we dive into the world of how A.I generated fake images are created using the popular Stable Diffusion LLM. Unlike Midjourney and Open AI's DALLE, Stable Diffusion's opensource nature has an extensive community and tools for custom training. But this hub for innovation has also sparked a wave of hyperrealistic A.I. fakes. In This video, we'll not just cover how these images are created but, more importantly, how to detect A.I. generated fake images.

Stable Diffusion Overview 0:25
Forms of A.I. Generated Fake Images 1:14
Why Stable Diffusion? 2:32
Stable Diffusion Framework for Fakes 3:16
Models and Checkpoints for Realistic Images 3:33
Training the Model (Text Embeddings and LoRa's) 4:18
Kohya_SS for Training 4:44
Using the LoRa in Stable Diffusion 5:11
User Created Model Database 5:40
How to Spot a Fake 6:12
Software Detection (AI Detectors) 9:17
Reverse Engineering a Photo 10:19
Other Fakes and Considerations 11:14