RATS Speaker Identification

<h3>Introduction</h3><br>
<p>RATS Speaker Identification was developed by LDC and is comprised of approximately 1,900 hours of Levantine Arabic, Farsi, Dari, Pashto and Urdu conversational telephone speech with annotations of speech segments. The audio was retransmitted over eight channels, making 17,000 hours of total audio. The corpus was created to provide training and development sets for the Speaker Identification (SID) task in the DARPA RATS (Robust Automatic Transcription of Speech) program.</p><br>
<p>The goal of the RATS program was to develop human language technology systems capable of performing speech detection, language identification, speaker identification and keyword spotting on the severely degraded audio signals that are typical of various radio communication channels, especially those employing various types of handheld portable transceiver systems. To support that goal, LDC assembled a system for the transmission, reception and digital capture of audio data that allowed a single source audio signal to be distributed and recorded over eight distinct transceiver configurations simultaneously. Those configurations included three frequencies -- high, very high and ultra high -- variously combined with amplitude modulation, frequency hopping spread spectrum, narrow-band frequency modulation, single-side-band or wide-band frequency modulation. Annotations on the clear source audio signal, e.g., time boundaries for the duration of speech activity, were projected onto the corresponding eight channels recorded from the radio receivers.</p><br>
<h3>Data</h3><br>
<p>The source audio consists of conversational telephone speech recordings collected by LDC specifically for the RATS program from Levantine Arabic, Pashto, Urdu, Farsi and Dari native speakers. Annotations on the audio files include start time, end time, speech activity detection (SAD) label, SAD provenance, speaker ID, speaker ID provenance, language ID, and language ID provenance.</p><br>
<p>The data is divided into training and development sets, each containing their own audio and annotation subdirectories.</p><br>
<p>All audio files are presented as single-channel, 16-bit PCM, 16000 samples per second; lossless FLAC compression is used on all files. When uncompressed, the files have typical "MS-WAV" (RIFF) file headers.</p><br>
<p>Annotation files are presented as tab-delimited, UTF-8 encoded, plain text.</p><br>
<h3>Sponsorship</h3><br>
<p>This material is based upon work supported by the Defense Advanced Research Projects Agency (DARPA) under Contract No. D10PC20016. The content does not necessarily reflect the position or the policy of the Government, and no official endorsement should be inferred.</p><br>
<h3>Samples</h3><br>
<p>Please view the following samples:</p><br>
<ul><br>
<li><a href="desc/addenda/LDC2021S08.flac">Source Audio Sample (FLAC)</a></li><br>
<li><a href="desc/addenda/LDC2021S08.tsv">Annotation Sample (TXT)</a></li><br>
<li><a href="desc/addenda/LDC2021S08.retran.flac">Retransmission Audio Sample (FLAC)</a></li><br>
<li><a href="desc/addenda/LDC2021S08.retran.tsv">Retransmission Annotation Sample (TXT)</a></li><br>
</ul><br>
<h3>Updates</h3><br>
<p>None at this time.</p></br>
Portions © 2014, 2015, 2017, 2018, 2021 Trustees of the University of Pennsylvania

### 引言
RATS说话人识别语料库由语言数据联盟（Linguistic Data Consortium，LDC）开发，包含约1900小时的黎凡特阿拉伯语、波斯语、达里语、普什图语与乌尔都语会话电话语音数据，并附带语音片段标注。原始音频经8个信道重传后，总音频时长达到17000小时。该语料库的构建目标是为美国国防高级研究计划局（Defense Advanced Research Projects Agency, DARPA）鲁棒语音自动转录（Robust Automatic Transcription of Speech, RATS）项目中的说话人识别（Speaker Identification, SID）任务提供训练集与开发集。

RATS项目的目标是研发能够在各类无线电通信信道（尤其是采用各类手持便携式收发机系统的信道）典型的严重退化音频信号上，完成语音检测、语言识别、说话人识别与关键词检索的人类语言技术系统。为支撑该目标，LDC搭建了一套音频数据传输、接收与数字化采集系统，可将单源音频信号同时通过8种不同的收发机配置进行分发与录制。这些配置涵盖高、甚高、超高三类频段，分别结合调幅、跳频扩频、窄带调频、单边带或宽带调频等调制方式。清晰源音频信号的标注（如语音活动的时间边界）会被投射到对应的8个由无线电接收机录制的信道上。

### 数据
源音频为LDC专为RATS项目从黎凡特阿拉伯语、普什图语、乌尔都语、波斯语与达里语母语者处采集的会话电话语音录音。音频文件的标注内容包括起始时间、结束时间、语音活动检测（Speech Activity Detection, SAD）标签、语音活动检测来源、说话人ID、说话人ID来源、语言ID以及语言ID来源。

该数据集划分为训练集与开发集，二者均包含独立的音频与标注子目录。

所有音频文件均采用单声道、16位脉冲编码调制（PCM）格式，采样率为16000样本/秒；所有文件均采用无损FLAC压缩。未压缩状态下，文件采用标准"MS-WAV"（RIFF）文件头格式。

标注文件采用制表符分隔、UTF-8编码的纯文本格式。

### 资助说明
本材料基于美国国防高级研究计划局（DARPA）合同D10PC20016资助的研究工作成果。本内容不一定反映美国政府的立场或政策，不应被视为获得官方认可。

### 示例
请查看以下示例：
- 源音频示例（FLAC）：desc/addenda/LDC2021S08.flac
- 标注示例（TXT）：desc/addenda/LDC2021S08.tsv
- 重传音频示例（FLAC）：desc/addenda/LDC2021S08.retran.flac
- 重传标注示例（TXT）：desc/addenda/LDC2021S08.retran.tsv

### 更新情况
暂无更新。

部分内容 © 2014、2015、2017、2018、2021 宾夕法尼亚大学理事会

搜集汇总

背景与挑战

背景概述

RATS Speaker Identification是一个用于说话人识别任务的数据集，包含约1,900小时电话对话语音，覆盖Levantine Arabic、Persian、Dari、Pushto和Urdu五种语言，音频经过8通道重传模拟退化信号。该数据集专为DARPA RATS项目设计，提供训练和开发集，支持语音活动检测、说话人ID和语言ID的标注，旨在提升在恶劣通信环境下的说话人识别技术。