1 related articles
An open-source workflow using LTX-2.3 and Face-ID LoRA that generates identity-locked talking videos from a single photo and voice recording. Supports CUDA and Apple Silicon locally.