A bell-foundry visitor desk can stream speech and drive a face (ACE / Riva / Audio2Face / NeMo LLM from official additional material) and can also follow a “listen, answer, smile” script. How should those concerns be split?
Select an answer to reveal the explanation.
Short Explanation
Microphones and faces are the ingredients; listen–answer–smile is the recipe card. Split modality orchestration (speech+face) from agent orchestration (the action script).
Full Explanation
In an avatar-style stack, speech and face animation are modalities to route and encode (modality orchestration), drawing on associate tool-selection among named pieces such as ACE, Riva, Audio2Face, and a NeMo LLM. The listen–answer–smile sequence is agent orchestration. Conflating both as fusion scores or OCR readiness is incorrect; swapping the labels also fails.