The First DDLP - Deep Dive Live Project!
This is where i have an idea and run it from start to finish, demonstrating a fully applied concept to completion. Along the way i can suggest branching choices which can be taken to further enhance the final version. This will use multiple services & methods with many AI services & software on display. In the end we can see what we must pay attention to while exposing some routes to improve this concept.
Links:
https://claude.ai/Civitai
https://civitai.com/
https://www.audacityteam.org/
https://www.kits.ai/
https://www.hedra.com/
https://adobe.com
https://lumalabs.ai/dream-machine
Sections:
00:00:00 Demo Video
00:00:30 Introduction to DDLP
00:05:30 Script Concept (CLAUDE)
00:14:08 Image Dataset (LORA)
00:17:52 Lora Training (CIVITAI)
00:21:28 Initial Image Gens (COMFYUI)
00:34:10 Puppet Options discussion
00:59:59 Speech Dataset (AUDACITY)
01:01:56 Phoneme Speech method
01:09:17 Speech Model Training (KITSAI)
01:15:22 Audacity, silence injection/removal
01:17:50 Lipsync Video gen (HEDRA)
01:24:31 Rotoscoping (AFTER EFFECTS)
01:31:09 Dream Machine BG scenes (LUMAAI)
01:42:05 First Draft
01:52:06 Results & Conclusions