0x3/Easy-Turn-Trainset
Easy Turn: Integrating Acoustic and Linguistic Modalities for Robust Turn-Taking in Full-Duplex Spoken Dialogue Systems Guojian Li1, Chengyou Wang1, Hongfei Xue1, Shuiyuan Wang1, Dehui Gao1, Zihan Zhang2, Yuke Lin2, Wenjie Li2, Longshuai Xiao2, Zhonghua Fu1,╀, Lei Xie1,╀ 1 Audio, Speech and Language Processing Group (ASLP@NPU), Northwestern Polytechnical University 2 Huawei Technologies, China 🎤 Demo Page 🤖 Easy Turn Model 📑 Paper 🌐 Huggingface… See the full description on the dataset page: https://huggingface.co/datasets/0x3/Easy-Turn-Trainset.
Easy Turn: Integrating Acoustic and Linguistic Modalities for Robust Turn-Taking in Full-Duplex Spoken Dialogue Systems
<p align="center"> Guojian Li<sup>1</sup>, Chengyou Wang<sup>1</sup>, Hongfei Xue<sup>1</sup>, Shuiyuan Wang<sup>1</sup>, Dehui Gao<sup>1</sup>, Zihan Zhang<sup>2</sup>, Yuke Lin<sup>2</sup>, Wenjie Li<sup>2</sup>, Longshuai Xiao<sup>2</sup>, Zhonghua Fu<sup>1</sup><sup>,╀</sup>, Lei Xie<sup>1</sup><sup>,╀</sup> </p>
<p align="center"> <sup>1</sup> Audio, Speech and Language Processing Group (ASLP@NPU), Northwestern Polytechnical University <br> <sup>2</sup> Huawei Technologies, China <br> </p>
<div align="center">
</div>
<p align="center"> <img src="src/logo.png" alt="Institution 5" style="width: 600px; border-radius: 30px;"> </p>
Download
The Easy Turn resources are available at Model, Trainset, and Testset.
Easy Turn Trainset
The Easy Turn Trainset is a large-scale audio dataset for turn-taking detection, comprising both real and synthetic data. It contains four subsets corresponding to different conversational turn-taking states: 580 hours of complete state, 532 hours of incomplete state, 10 hours of backchannel state, and 23 hours of wait state, totaling approximately 1,100 hours. Each recording is accompanied by a text transcription and labeled with one of the four turn-taking states. <div align="center"><img width="550px" src="src/data_pipeline.jpg" /></div>
Citation
Please cite our paper if you find this work useful: