TIMIT Acoustic-Phonetic Continuous Speech Corpus
收藏资源简介:
<h3>Introduction</h3><br> <p>The TIMIT corpus of read speech is designed to provide speech data for acoustic-phonetic studies and for the development and evaluation of automatic speech recognition systems. TIMIT contains broadband recordings of 630 speakers of eight major dialects of American English, each reading ten phonetically rich sentences. The TIMIT corpus includes time-aligned orthographic, phonetic and word transcriptions as well as a 16-bit, 16kHz speech waveform file for each utterance. Corpus design was a joint effort among the Massachusetts Institute of Technology (MIT), SRI International (SRI) and Texas Instruments, Inc. (TI). The speech was recorded at TI, transcribed at MIT and verified and prepared for CD-ROM production by the National Institute of Standards and Technology (NIST).</p><br> <p>The TIMIT corpus transcriptions have been hand verified. Test and training subsets, balanced for phonetic and dialectal coverage, are specified. Tabular computer-searchable information is included as well as written documentation.</p><br> <h3>Samples</h3><br> <ul><br> <li><a href="desc/addenda/LDC93S1.phn" rel="nofollow">phonemes</a></li><br> <li><a href="desc/addenda/LDC93S1.txt" rel="nofollow">transcripts</a></li><br> <li><a href="desc/addenda/LDC93S1.wav" rel="nofollow">audio</a></li><br> <li><a href="desc/addenda/LDC93S1.wrd" rel="nofollow">word list</a></li><br> </ul></br> Portions © 1993 Trustees of the University of Pennsylvania
<h3>引言</h3><br><p>TIMIT朗读语音语料库旨在为声学语音学研究以及自动语音识别(Automatic Speech Recognition)系统的开发与评估提供语音数据。该语料库收录了8种主要美式英语方言的630名说话者的宽带语音录音,每位说话者朗读10条富含语音学信息的句子。TIMIT语料库为每一条语音话语均提供时间对齐的正字法转写、音素转写与单词转写,以及16位、16kHz的语音波形文件。本语料库的设计由麻省理工学院(Massachusetts Institute of Technology, MIT)、国际SRI公司(SRI International)与德州仪器公司(Texas Instruments, Inc., TI)联合完成,语音数据由TI录制、MIT负责转写,美国国家标准与技术研究院(National Institute of Standards and Technology, NIST)则承担数据验证与CD-ROM出版筹备工作。</p><br><p>TIMIT语料库的所有转写均经过人工校验。语料库已划定针对语音覆盖与方言覆盖均均衡的测试集与训练子集,除书面文档外,还提供可供计算机检索的表格化数据。</p><br><h3>示例数据</h3><br><ul><br><li><a href="desc/addenda/LDC93S1.phn" rel="nofollow">音素文件</a></li><br><li><a href="desc/addenda/LDC93S1.txt" rel="nofollow">转写文本</a></li><br><li><a href="desc/addenda/LDC93S1.wav" rel="nofollow">音频文件</a></li><br><li><a href="desc/addenda/LDC93S1.wrd" rel="nofollow">单词列表</a></li><br></ul></br>部分内容 © 1993 宾夕法尼亚大学托管委员会

- TIMIT Acoustic-Phonetic Continuous Speech Corpus首次发表,由美国国防部高级研究计划局(DARPA)资助,旨在为语音识别研究提供标准化的语音数据集。
- TIMIT数据集首次应用于语音识别研究,成为语音处理领域的重要基准数据集,促进了语音识别技术的发展。
- 随着计算能力的提升和深度学习技术的兴起,TIMIT数据集被广泛用于训练和评估语音识别模型,进一步推动了语音处理技术的进步。
- TIMIT数据集在语音识别领域的应用扩展到多语言和跨语言研究,为全球语音处理技术的多样化和国际化提供了重要支持。
- TIMIT数据集继续作为语音处理研究的基础数据集,支持了包括语音合成、语音增强和语音情感识别在内的多个前沿研究方向。
- 1The DARPA TIMIT Acoustic-Phonetic Continuous Speech CorpusTexas Instruments, Massachusetts Institute of Technology · 1990年
- 2Deep Neural Networks for Acoustic Modeling in Speech Recognition: The Shared Views of Four Research GroupsIEEE · 2012年
- 3Investigation of Full-Length Speech Recognition Using Deep Neural Network Acoustic ModelsIEEE · 2013年
- 4Phoneme Recognition on the TIMIT DatabaseSpringer · 2013年
- 5A Study on Speaker Adaptation of the Parameters of Continuous Density Hidden Markov Models Using the TIMIT DatabaseIEEE · 1992年



