From Turn-Taking to Synchronous Dialogue: A Survey of Full-Duplex Spoken Language Models

Yuxuan Chen, Haoyuan Yu
9/18/2025
cs.CLcs.SDeess.AS

No official code found. Best implementation: elpsykongloo/FD-SLMs (34 stars, last commit 2025-12-23).

Abstract

True Full-Duplex (TFD) voice communication--enabling simultaneous listening and speaking with natural turn-taking, overlapping speech, and interruptions--represents a critical milestone toward human-like AI interaction. This survey comprehensively reviews Full-Duplex Spoken Language Models (FD-SLMs) in the LLM era. We establish a taxonomy distinguishing Engineered Synchronization (modular architectures) from Learned Synchronization (end-to-end architectures), and unify fragmented evaluation approaches into a framework encompassing Temporal Dynamics, Behavioral Arbitration, Semantic Coherence, and Acoustic Performance. Through comparative analysis of mainstream FD-SLMs, we identify fundamental challenges: synchronous data scarcity, architectural divergence, and evaluation gaps, providing a roadmap for advancing human-AI communication.

View on arXivView PDF

Code Implementations(1)

elpsykongloo/FD-SLMsCommunity67%

This is an evolving repo for the paper “From Turn-Taking to Synchronous Dialogue: A Survey of Full-Duplex Spoken Language Models ”A comprehensive survey of Full-Duplex Spoken Language Models (FD-SLMs) -- For ICASSP 2026.

340Sep 17, 202510 months ago

Similar Papers

Beyond Turn-Based Interfaces: Synchronous LLMs as Full-Duplex Dialogue Agents

89% match2024

Full-Duplex-Bench: A Benchmark to Evaluate Full-duplex Spoken Dialogue Models on Turn-taking Capabilities

88% match1 repo3142025

Language Model Can Listen While Speaking

88% match2024

A Full-duplex Speech Dialogue Scheme Based On Large Language Models

86% match2026

MTR-DuplexBench: Towards a Comprehensive Evaluation of Multi-Round Conversations for Full-Duplex Speech Language Models

86% match2025

FLEXI: Benchmarking Full-duplex Human-LLM Speech Interaction

85% match2025

Cite this paper

@article{chen2025from,
  title  = {From Turn-Taking to Synchronous Dialogue: A Survey of Full-Duplex Spoken Language Models},
  author = {Yuxuan Chen and Haoyuan Yu},
  year   = {2025},
  eprint = {2509.14515},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2509.14515v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2509.14515.svg)](https://codeofpaper.com/paper/2509.14515)

Discussion