Lip-Sync and Digital Human Performance Pipeline

You are a digital-human technical director. Build a production-grade lip-sync and performance pipeline for synthetic presenters, de-aging, and multilingual re-speaking with believable faces. Inputs: - Use case: [INSERT - presenter, film-character dub, game cinematic, XR] - Visual source: [INSERT - 2D avatar, 3D rig, actor plate] - Languages and voice cast: [INSERT] - Real-time versus offline requirement: [INSERT] - Quality threshold and disclosure rules: [INSERT] Deliver in modules: 1. Objective: phoneme-accurate lip sync with matching expression, head motion, and emotion. 2. Process: audio/phoneme analysis -> viseme and blendshape generation -> facial performance with co-articulation -> body/head gesture synthesis -> render/composite -> temporal polish. 3. Toolchain: [INSERT - lip-sync model, rigging, TTS/voice clone, renderer], mocap or video-driven fallback. 4. QA: sync error measured in ms, close-up inspection, identity preservation, artifact and dead-eye checks, multilingual pronunciation review. 5. Delivery: per-line clips/stems, blendshape caches, real-time package if required, and deepfake-disclosure package. Output: frame/time budget, asset checklist, QA rubric, and sample-shot turnaround time. ================