Abstract:Spoken dialogue models have significantly advanced intelligent human-computer interaction, yet they lack a plug-and-play full-duplex prediction module for semantic endpoint detection, hindering seamless audio interactions. In this paper, we introduce Phoenix-VAD, an LLM-based model that enables streaming semantic endpoint detection. Specifically, Phoenix-VAD leverages the semantic comprehension capability of the LLM and a sliding window training strategy to achieve reliable semantic endpoint detection while supporting streaming inference. Experiments on both semantically complete and incomplete speech scenarios indicate that Phoenix-VAD achieves excellent and competitive performance. Furthermore, this design enables the full-duplex prediction module to be optimized independently of the dialogue model, providing more reliable and flexible support for next-generation human-computer interaction.
From: Weijie Wu [view email]
[v1]
Wed, 24 Sep 2025 07:09:19 UTC (1,682 KB)
[v2]
Fri, 26 Sep 2025 03:37:38 UTC (1,682 KB)
[v3]
Thu, 30 Oct 2025 06:30:08 UTC (1 KB) (withdrawn)
[v4]
Tue, 4 Nov 2025 07:04:52 UTC (1,682 KB)
[v5]
Thu, 3 Sep 2026 03:59:24 UTC (1,982 KB)
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。