CATT: Complexity-Aware Thinker–Talker Orchestration for Low-Latency Spoken Responses in Digital Humans
Electronics · 01.11.2026 · 15 (21) · 5064
DOI: 10.3390/electronics15215064 ↗Kurz gefasst
Fachartikel in der Zeitschrift Electronics, online erschienen am 01.11.2026. Verfasst von Xi Cheng und 4 weiteren Autor:innen. Die Arbeit nennt 61 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Digital humans must speak promptly, but thinking-enabled large language models (LLMs) answer only after reasoning. CATT (Complexity-Aware Thinker–Talker) routes easy turns to a direct reply; hard turns get an answer-free acknowledgment, progress speech from background reasoning, then the final answer. It needs no fine-tuning and is plug-and-play: its orchestration code runs unchanged when the model endpoint, router, or speech and playback layer is replaced. Evaluated on the transcript-to-speech subsystem (local Qwen3-32B, 738 isolated turns, and virtual playback), CATT starts speaking about 6.3 s earlier than a prompt-matched thinking baseline. The first substantive content comes about 10.5 s earlier on closed-form math and science questions and 3.0 s earlier on open-ended requests (about 1 s later when routed to reasoning). The final closed-form answer arrives about 3.2 s later at the same accuracy. An LLM-judged listening rating, a proxy that includes waiting, places CATT about 1 point above the thinking baseline but 0.42 below a non-thinking baseline on open-ended requests. Both latency gains held for three model families accessed through hosted APIs and in an end-to-end avatar case study. A limited ReVerT reproduction on the same reasoning started speaking 8.4 s after CATT. Real users, microphones, networks and multi-user load remain untested.
Zitieren
APA
Cheng, X., Ding, Z., Xu, F., Xu, K., & Chen, T. (2026). CATT: Complexity-Aware Thinker–Talker Orchestration for Low-Latency Spoken Responses in Digital Humans. Electronics, 15(21), 5064. https://doi.org/10.3390/electronics15215064
Vancouver
Cheng X, Ding Z, Xu F, Xu K, Chen T. CATT: Complexity-Aware Thinker–Talker Orchestration for Low-Latency Spoken Responses in Digital Humans. Electronics. 2026;15(21):5064. doi:10.3390/electronics15215064
BibTeX
@article{Cheng2026,
author = {Cheng, Xi and Ding, Zezhen and Xu, Fengyu and Xu, Kaidi and Chen, Tianlong},
title = {CATT: Complexity-Aware Thinker–Talker Orchestration for Low-Latency Spoken Responses in Digital Humans},
journal = {Electronics},
year = {2026},
volume = {15},
number = {21},
pages = {5064},
doi = {10.3390/electronics15215064}
}
EUVIMED