Guest
Complite guide on how to use Stable Diffusion with Google Colab Notebook to start AI animation from your photo and make cool AI video transition that you saw trending on TikTok. I explained the workflow on two examples with dungeon and photorealistic models. No need to install SD on your local drive. Analog Diffusion (photorealistic model): prompts: DUNGEONS N Waifu's model: SD Deforum 0.7 (default): Quick Guide to Deforum v06: Deforum Gradient Conditioning Spreadsheet: #gid=0 If you face any problem - go to Deforum Discord server: Marta Bula Dancer: Unisex t-shirts: The Music you hear I also produce: Qaqarotto - Never Again Timecode: 00:00 - Examples what you will learn today 00:59 - Preparation: idea , export frame from video or take photo 02:42 - Google Drive 03:00 - ANALOG DIFFUSION model (photorealistic) 03:31 - Google Colab Notebook 04:08 - Model Setup 05:23 - Animation Settings (3D) 06:49 - Prompts (ANALOG DIFFUSION) 07:42 - Image aand Sampling 08:17 - seed_behavior 08:49 - Input Settings (your photo here) 09:36 - Create video without any Video Editor 10:21 - Davinci Resolve (noise reduction, deflicker, LUT) 12:11 - DUNGEONS N Waifu's model (trending on TikTok) 13:06 - Google Colab notebook parametrs 13:56 - Color Coherence 14:10 - Prompts (DUNGEONS) 15:08 - Image, Sampling, Batch, Init 15:40 - Exposure/Contrast Conditional Settings 16:24 - Last words
