[论文] Multi-Faceted Interactivity Alignment in Full-Duplex Speech Models

论文概要

研究领域: NLP 作者: Atsumoto Ohashi, Neil Zeghidour, Alexandre Défossez, Eugene Kharitonov 发布时间: 2026-06-09 arXiv: 2606.11167

中文摘要

全双工语音对话模型可同时听和说，但现有模型仅用监督学习通过token级似然最大化训练，不直接优化交互级行为，导致过度沉默和话轮转换时机不当等问题。本文提出后训练对齐方法，通过RL全面改善交互性：处理停顿、话轮转换、回馈语和用户打断四个维度。每个维度从人类对话语料提取短音频片段，用特定奖励函数优化。应用于Moshi和PersonaPlex，在离线评估和实时多轮对话评估中均实现一致提升。

原文摘要

Full-duplex spoken dialogue models can listen and speak simultaneously, making them a promising architecture for natural conversation. However, current models are trained solely with supervised learning through token-level likelihood maximization, which does not directly optimize interaction-level behaviors, causing interactivity issues such as excessive silence and ill-timed turn-taking. Recent work has applied reinforcement learning (RL) to improve interactivity, but existing methods address only a limited set of interactive behaviors in their rewards. In this work, we propose a post-training alignment method that comprehensively improves the interactivity of full-duplex spoken dialogue models through RL. We address the four canonical axes of interactivity: pause handling, turn-taking, b...

--- *自动采集于 2026-06-11*

#论文 #arXiv #NLP #小凯