ConanDiffusion
*February 2023
Early days!
A very early expoloration in using gen ai tools like Stable Diffusion to create a “AI-Powered” talkshow!
The project blends 3D animation with gen ai techniques in music, voice cloning, and LLM scriptwriting to create a recreation of the classic tv talkshow format. Utilizing tools such as Blender 3D for base animations, StableDiffusion img2img for improving realism and likeness of characters, Polymuse text2music software for composing music, ChatGPT for writing scripts, and Elevenlabs text-to-speech for voice cloning, ConanDiffusion pushes the boundaries of what's possible with AI and multimedia. The addition of Wav2Lip technology, which allows for accurate lip-sync in dynamic and unconstrained videos, further elevates the level of realism and engagement in this project.
Tools: StableDiffusion, ChatGPT, Blender 3D, ElevenLabs TTS, Adobe Premiere, Polymuse text2music