arXiv：cs.AI· Cheng-Kuang Chang, Alexander H. Liu, Brian Kingsbury, Hilde Kuehne, Rogerio Feris, Samuel Thomas, James Glass, Kai-Wei Chang·· 22 小时前
D2：从半双工骨干模型转换的高能力全双工语音语言模型
D2: High-Capability Full-Duplex Spoken Language Models from Half-Duplex Backbones
AI 导读
研究者提出 Dialogue-to-Duplex（D2）框架，可将预训练半双工语音语言模型（HD-SLM）转换为高能力全双工模型（FD-SLM），并把该方法应用于 Qwen3-Omni 和 LLaMA-Omni 2 两种架构。
正文
Abstract:Full-duplex spoken language models (FD-SLMs) support simultaneous listening and speaking, but often lag behind half-duplex spoken language models (HD-SLMs) in general capabilities. We propose Dialogue-to-Duplex (D2), a framework that converts pretrained HD-SLMs into high-capability FD-SLMs at fine temporal granularity. D2 adapts the original audio encoder to streaming through decoder-free representation distillation and learns full-duplex interaction through supervised fine-tuning. We systematically study the effects of chunk size on full-duplex interaction, general capabilities, and inference computation. We apply D2 to two distinct HD-SLM architectures, Qwen3-Omni and LLaMA-Omni 2. At a 12.5Hz update rate, Qwen3-Omni-D2 achieves 99.2% turn-taking success with 521-ms latency, 99.5% interruption success, and a state-of-the-art VoiceBench score of 70.1.
| Comments: | 17 pages, including references and appendices. Demo: this https URL |
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Audio and Speech Processing (eess.AS) |
| Cite as: | arXiv:2610.17124 [cs.CL] |
| (or arXiv:2610.17124v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.17124 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Cheng Kuang Chang [view email]
[v1]
Mon, 12 Oct 2026 17:15:29 UTC (1,630 KB)
来源：arXiv：cs.AI · arxiv.org