RSL2019: A Realistic Speech Localization Corpus

We present a new database for speech localization that we refer to as Realistic Speech Localization 2019 (RSL2019) corpus. The corpus is designed for the study of sound source localization in real-world applications. The RSL2019 corpus is a continuing effort, which presently contains 22.60 hours of speech data, recorded using a four channel microphone array, and played over a loudspeaker from different directions of arrival (DOA). We consider 180speech utterances spoken by 6 speakers, selected from RSR2015database, which are played over the loudspeaker positioned at different angles and distances from the microphone array. We vary the DOA from 0 to 360 degree angle at an interval of 5degree, at 1 metre and 1.5 metre distance. From each position and DOA, we also record white noise to study the robustness, and time stretched pulse to generate the transfer function for speech localization algorithm. Furthermore, we present the experimental results and analysis on state-of-the-art sound source localization algorithm using the open source HARK toolkit on the created RSL2019 database. This database is provided for research purpose only. If you use this database, please cite the following paper. Rohan Sheelvant, Bidisha Sharma, Maulik Madhavi, Rohan Kumar Das, S.R.M. Prasanna and Haizhou Li, "RSL2019: A Realistic Speech Localization Corpus," <em>2019 22nd Conference of the Oriental COCOSDA International Committee for the Co-ordination and Standardisation of Speech Databases and Assessment Techniques (O-COCOSDA)</em>, 2019, pp. 1-6, doi: 10.1109/O-COCOSDA46868.2019.9060842. .本研究提出一款全新的语音定位数据库，命名为Realistic Speech Localization 2019（RSL2019）语料库。该语料库专为真实场景下的声源定位研究设计。RSL2019语料库为持续迭代的研究资源，当前包含22.60小时的语音数据：采用四通道麦克风阵列进行录制，语音信号通过扬声器播放，并设置不同的到达方向（Direction of Arrival, DOA）。本语料库选用来自RSR2015数据库的6名说话者的180条语音片段，将其通过置于距麦克风阵列不同角度与距离位置的扬声器播放。我们将到达方向（DOA）以5度为间隔在0至360度范围内调整，并设置1米与1.5米两种播放距离。针对每一组位置与到达方向参数，我们额外录制了白噪声以测试算法鲁棒性，同时录制时伸脉冲（time-stretched pulse）以生成语音定位算法所需的传递函数。此外，本研究基于构建的RSL2019数据库，采用开源HARK工具包对当前前沿的声源定位算法开展了实验并给出结果与分析。本数据库仅用于学术研究用途。若您使用该数据库，请引用以下文献：Rohan Sheelvant、Bidisha Sharma、Maulik Madhavi、Rohan Kumar Das、S.R.M. Prasanna及Haizhou Li，"RSL2019: A Realistic Speech Localization Corpus"，发表于2019年第22届东方COCOSDA国际语音数据库与评估技术协调标准化委员会（Oriental COCOSDA International Committee for the Co-ordination and Standardisation of Speech Databases and Assessment Techniques, O-COCOSDA）会议，2019年，第1-6页，DOI: 10.1109/O-COCOSDA46868.2019.9060842。