Hi everyone, I'm a third-year engineering student working on an AI project that converts speech into a fully animated 3D character. Right now, my prototype works like this: Audio ↓ EMAGE (Speech → Motion) ↓ SMPL Motion (Research Prototype) ↓ Motion Retargeting (planned) ↓ Commercial Character Rig ↓ Blender Rendering ↓ MP4 Animation The current implementation uses SMPL only as an internal research representation . Since SMPL isn't intended for unrestricted commercial use, my long-term...