TurnBench is a new multi‑domain benchmark for evaluating turn‑taking dynamics in spoken dialogue. It comprises a 30‑hour hand‑labeled corpus of dyadic human conversations, a standardized evaluation protocol for end‑of‑turn and interruption detection, and covers six distinct interaction styles with triple annotation. The benchmark also provides a 104‑hour training set, a public leaderboard, and an interactive dataset viewer at https://turnbench.sesame.com.
By Freeman Jiang, Ramon Sanabria, Soham Deshmukh, Bandhav Veluri, Simon Michael Vuch Williams, Elliott K. Suen, Garreth Lee, Kevin Yoonho Choi, Takuya Umeki, Riku Kubo, Sathvik Udupa, Chien-yu Huang, Shih-Yun Shan Kuan, Zhuoyan Tao, Satyapriya Krishna, Sefik Emre Eskimez, Yu Tsao, Hung-yi Lee, Shinji Watanabe
ECHO is a new paired diagnostic benchmark designed for Chinese full‑duplex spoken dialogue systems to evaluate context‑sensitive turn‑taking. It pairs examples that share the same overlap transcript but differ in preceding multi‑turn context, requiring either a Yield or Keep decision, and also includes off‑talk cases to test unnecessary yielding. The benchmark introduces pair accuracy, penalizing constant‑action policies, and shows that many systems are biased toward Yield, performing better on interruptions than on backchannels.
By Shuofeng Zhao, Hongwei Cai, Wenke Fan, Qingxiang Guo, Dawei Yang, Zhou Wang, Zhiyang Zhou, Yingxin Shang, Weixu Wang, Lin Yang, Shuran Zhou, Yang Song
DuplexSpeechBench-IFEval (DSB-IFEval) is a new benchmark that evaluates how full‑duplex voice agents follow implicit instructions during real‑time spoken interaction. It contains 1,038 test cases across eight assistant roles and tests five conditioning protocols, measuring floor management with an Instruction Adherence Score (IAS) and persona consistency with a Persona Adherence Score (PAS). Experiments on six speech systems reveal architecture‑dependent trade‑offs, showing that some models are more sensitive to explicit versus persona‑only instructions and that even when following conflicting directives, they struggle to override them under safety conflict.
By Puneet Mathur, Dinesh Manocha
arXiv:2609.13117v1 Announce Type: new
Abstract: Full-duplex evaluation often emphasizes whether an agent keeps speaking or stops. That binary cannot express a third response humans use routinely: con...
By Yunqi Lu, Tyler Baumgartner, Nikhil Johri, Brandon Tai, Candice Fan, Luc Debaupte, Ruben Aguilar, Bill Wang, Yi Zhong
arXiv:2609.13076v1 Announce Type: cross
Abstract: Conversational voice agents have advanced significantly, offering increasingly natural human-machine interactions through both cascaded and end-to-en...
By Yi-Jen Shih, Shih-Yun Shan Kuan, Guan-Ting Lin, Kai-Wei Chang, Siddhant Arora, Shu-wen Yang, Abdelrahman Mohamed, Shinji Watanabe, Hung-yi Lee, David Harwath
The paper introduces a pipeline that generates intent‑labeled, two‑channel conversational speech from relational event lists, enabling controlled synthesis of full‑duplex dialogue with 42 phenomena across eight families in English and Mandarin. By having an LLM author each event’s speaker, text, conversational act, and attachment, and then aligning and timing these events independently, the system produces diverse, realistic turn‑taking signals. Experiments show that models trained on this synthetic corpus achieve higher floor‑occupancy accuracy and better start‑speaking/listening F1 scores compared to models trained on prior data.
By Matthew Sun, Vinay Kothapally, Meng Yu, Chao Huang, Hao Zhang, Yixuan Zhang, Steve Yves
The study evaluates conversational speech models by placing them at 427 pauses and 482 turn ends from 176 unscripted conversations, scoring their responses on timing, latency, length, wording, and recall of recent dialogue. Hosted APIs (GPT Realtime, Gemini 3.1 Flash Live, Grok Voice) respond at almost every pause, with latency ranging from 1–4 s for silence-based endpointers and about 8 s for semantic turn detection, while full‑duplex models (PersonaPlex, Moshi) remain silent more often and can interrupt within human reaction time. Recall rates vary: hosted systems remember 41–60 % of facts under a Claude Opus 5.5 judge and 19–26 % with a strict word matcher; full‑duplex models recall 17–22 % and 12–15 % respectively.
By Arya Bastani, Iman Modarressi
arXiv:2610.08125v1 Announce Type: cross
Abstract: Full-duplex spoken dialogue models listen and speak at the same time, enabling voice agents to have natural, low-latency interactions that turn-based...
By Sungnyun Kim, Sungwoo Cho, Jihwan Oh, Se-Young Yun
arXiv:2609.27372v1 Announce Type: cross
Abstract: Benchmarks for full-duplex spoken dialogue models score turn-taking with binary fixed-window rules that reward immediate response or silence by compl...
By Kian Shamsaie, Iman Modarressi
arXiv:2609.20995v1 Announce Type: cross
Abstract: Natural spoken interaction requires more than streaming ASR, language generation, and speech synthesis: a system must react to overlap without cancel...
By Bertil Braun
SteerDuplex is a full‑duplex speech dialogue model that can be steered along attributes such as tone, persona, speaking rate, and voice style in response to user instructions. The authors introduce a taxonomy of text‑ and audio‑based steerability, identify gaps in existing models, and fine‑tune a Moshi‑based model with reinforcement learning to improve timing and response continuity. They also present SteerBench, a benchmark of 390 spoken prompts and 1,067 human‑authored rubrics, showing significant gains in audio‑steering pass rates and interruption handling compared to open baselines.
By Utkarsh Tyagi, Ramaneswaran Selvakumar, Advait Gosai, Sonal Kumar, Nikhil Barhate, Isabell Sagar, Steven Li, Miheer Bavare, Daniel Quigley, Fabiola Tapia Carrillo, Jose M Patron E, Diego Mac\'ias Guti\'errez, Paul Song, Ramani Duraiswami, Dinesh Manocha, Yunzhong He
The paper introduces ECHO, a paired diagnostic benchmark for evaluating Chinese real‑time spoken dialogue systems on turn‑taking. ECHO pairs examples that share the same overlap transcript but differ in preceding multi‑turn context, requiring either Yield or Keep actions, and also includes off‑talk cases to test unnecessary yielding. Experiments on four speech systems reveal that three systems over‑yield, correctly keeping the floor on fewer than 13% of backchannels, while the fourth system shows a more balanced performance, illustrating that interruption‑only evaluation can overestimate turn‑taking reliability.
By Shuofeng Zhao, Hongwei Cai, Wenke Fan, Qingxiang Guo, Zhou Wang, Dawei Yang, Zhiyang Zhou, Yingxin Shang, Weixu Wang, Lin Yang, Shuran Zhou, Yang Song