Multi-Faceted Interactivity Alignment in Full-Duplex Speech Models

• URL: http://arxiv.org/abs/2606.11167v1
• Date: Tue, 09 Jun 2026 17:46:55 GMT
• Title: Multi-Faceted Interactivity Alignment in Full-Duplex Speech Models
• Abstract summary: We propose a post-training alignment method that improves full spoken dialogue models through RL.<n>We address the four canonical axes of interactivity: handling turn-taking, back-channel pause, and user interruption.<n>An extra-based reward for response quality prevents semantic degradation.
• Score: 53.470209949659115
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: 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, backchanneling, and user interruption. For each axis, we extract short audio segments from human conversation corpora and optimize the model with axis-specific reward functions. An extra LLM-based reward for response quality prevents semantic degradation. We apply our method to two open-source models, Moshi and PersonaPlex, demonstrating consistent improvements in interactivity on both offline evaluation with pre-recorded audio and real-time multi-turn dialogue evaluation.

Related papers

• ASPIRin: Action Space Projection for Interactivity-Optimized Reinforcement Learning in Full-Duplex Speech Language Models [49.18708573170585]
  Empirical evaluations show ASPIRin optimize interactivity across turntaking, backchanneling, and pause handling.<n>We propose ASPIRin degrade turn to say when when to collapse when to say.
  arXiv  Detail & Related papers  (2026-04-11T07:07:08Z)
• FLEXI: Benchmarking Full-duplex Human-LLM Speech Interaction [49.83226596963294]
  Speech-computer human interaction enables real-time spoken dialogue systems.<n>Modelling and benchmarking these models remains a fundamental challenge.<n>We introduce FLEXI, the first benchmark for full-human spoken interaction.
  arXiv  Detail & Related papers  (2025-09-26T11:57:42Z)
• Aligning Spoken Dialogue Models from User Interactions [55.192134724622235]
  We propose a novel preference alignment framework to improve spoken dialogue models on realtime conversations from user interactions.<n>We create a dataset of more than 150,000 preference pairs from raw multi-turn speech conversations annotated with AI feedback.<n>Our findings shed light on the importance of a well-calibrated balance among various dynamics, crucial for natural real-time speech dialogue systems.
  arXiv  Detail & Related papers  (2025-06-26T16:45:20Z)
• Full-Duplex-Bench: A Benchmark to Evaluate Full-duplex Spoken Dialogue Models on Turn-taking Capabilities [93.09944267871163]
  FullDuplexBench is a benchmark that systematically evaluates key interactive behaviors.<n>By releasing our benchmark code we aim to advance spoken dialogue modeling and the development of more natural and engaging SDMs.
  arXiv  Detail & Related papers  (2025-03-06T18:59:16Z)
• OmniFlatten: An End-to-end GPT Model for Seamless Voice Conversation [53.7173034249361]
  End-to-end GPT-based model OmniFlatten capable of effectively modeling complex behaviors inherent natural conversations with low latency.<n>Our approach offers a simple modeling technique and a promising research direction for developing efficient and natural end-to-end full- spoken dialogue systems.
  arXiv  Detail & Related papers  (2024-10-23T11:58:58Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.