Find research datasets worth reusing
Search datasets from major research repositories and use ShareScore to quickly assess how well each record supports discovery, access, and reuse.
21
datasets available to search
ShareScore release 0.9.0
Dataset results
21 results for “singing dataset”
N20EMv2 dataset for automatic music transcription from multimodal singing
<p>N20EMv2 dataset for multimodal automatic music transcription from multimodal singing, presented in our TOMM 2024 paper, Automatic Lyric Transcription and Automatic Music Transcription from Multimodal Singing. This dataset contains recordings of two modalities: audio and video. </p> <p>Our paper is available at: https://dl.acm.org/doi/10.1145/3651310.</p> <p>Code is available at: https://github.com/guxm2021/SVT_SpeechBrain</p> <p>Please cite our work as:</p> <pre>@article{gu2024automatic, title={Automatic Lyric Transcription and Automatic Music Transcription from Multimodal Singing}, author={Gu, Xiangming and Ou, Longshen and Zeng, Wei and Zhang, Jianan and Wong, Nicholas and Wang, Ye}, journal={ACM Transactions on Multimedia Computing, Communications and Applications}, publisher={ACM New York, NY}, year={2024} }</pre>
Jingju a cappella singing dataset part1
<p>This is the 4th version of the dataset. The folder structure has been changed since the 2nd version, where the Laosheng folder has been moved directly into wav or textgrid folder.</p> <p><strong>Description:</strong></p> <p>This dataset is a collection of boundary annotations of a cappella singing performed by Beijing Opera (Jingju, 京剧) professional and amateur singers. </p> <ol> <li>wav.zip: audio files in .wav format, mono or stereo.</li> <li>wav_mono.zip: audio files in .wav format, mono</li> <li>pycode.zip: util code for parsing the .textgrid annotation</li> <li>catalogue*.csv: recording metadata, source separation recordings are not included.</li> <li>textgrid.zip: phrase, syllable and phoneme annotation in Praat .textgrid format</li> <li>annotation_txt.zip: phrase, syllable and phoneme time boundaries (second) and labels in .txt format <ol> <li>*phrase_char: phrase-level time boundaries, labeled in Mandarin characters</li> <li>*phrase: phrase-level time boundaries, labeled in Mandarin pinyin</li> <li>*syllable: syllable-level time boundaries, labeled in Mandarin pinyin</li> <li>*phoneme: phoneme-level time boundaries, labeled in X-SAMPA</li> </ol> </li> </ol> <p>The boundaries (onset and offset) have been annotated in both <strong>Praat TextGrid (textgrid.zip)</strong> and .<strong>txt (annotation_txt.zip)</strong> format hierarchically:</p> <ol> <li>phrase (line),</li> <li>syllable,</li> <li>phoneme</li> </ol> <p>Singing units in pinyin and X-SAMPA have been annotated to a jingju a cappella singing audio dataset.</p> <p>The corresponding audio files are the a cappella singing arias recordings, which are stereo or mono, sampled at 44.1 kHz, and stored as .wav files. The .wav files are recorded by two institutes: those file names ending with ‘qm’ are recorded by C4DM, Queen Mary University of London; others file names ending with ‘upf’ or ‘lon’ are recorded by MTG-UPF. Additionally, another collection of 15 clean singing recordings is included in this dataset. They are extracted from the commercial recordings which originally contains karaoke accompaniment and mixed versions.</p> <p><strong>If you use this audio dataset in your work, please cite (1) this dataset as well (2) the following publication:</strong></p> <blockquote> <p>D. A. A. Black, M. Li, and M. Tian, “Automatic Identification of Emotional Cues in Chinese Opera Singing,” in 13th Int. Conf. on Music Perception and Cognition (ICMPC-2014), 2014, pp. 250–255.</p> </blockquote> <p> </p> <p><strong>Details:</strong><br> Annotation format, units, parsing code and other information please refer to <a href="https://github.com/MTG/jingjuPhonemeAnnotation">https://github.com/MTG/jingjuPhonemeAnnotation</a></p> <p><br> <strong>License:</strong><br> Textgrid annotations are licensed under Creative Commons Attribution-NonCommercial 4.0 International License.</p> <p>Wav audio ending with ‘upf’ or ‘lon’ is licensed under Creative Commons Attribution-NonCommercial 4.0 International.</p> <p>For the license of .wav audio ending with ‘qm’ from C4DM Queen Mary University of London, please refer to this page <a href="http://isophonics.org/SingingVoiceDataset">http://isophonics.org/SingingVoiceDataset</a></p> <p><strong>Contact information:</strong></p> <p>Rong Gong: rong<dot>gong<at>upf<dot>edu</p> <p>Rafael Caro Repetto: rafael<dot>caro<at>upf<dot>edu</p>
Jingju a cappella singing syllable boundary and duration annotation dataset
<p>This dataset is a collection of syllable boundary annotations and syllable duration annotations of a cappella singing performed by jingju (京剧, Beijing opera) professional and amateur singers. This dataset was used as the experimental dataset in the following work:</p> <blockquote> <p>Rong Gong, Nicolas Obin, Georgi Dzhambazov and Xavier Serra, “Score-Informed syllable segmentation for jingju a cappella singing voice with Mel-frequency intensity profiles," in<em> Folk Music Analysis workshop (FMA) 2017, Málaga, Spain</em></p> </blockquote> <p><strong>Audio Content</strong></p> <p>The audio files are the a cappella singing arias recordings, which are stereo or mono, sampled at 44.1 kHz, and stored as wav files. They can be found at this link http://doi.org/10.5281/zenodo.344932</p> <p>The wav files are recorded by two institutes: those file names ending with ‘qm’ are recorded by C4DM Queen Mary University of London; others file names ending with ‘upf’ or ‘lon’ are recorded by MTG-UPF. If you use the dataset in your work, please cite the following publication.</p> <blockquote> <p>D. A. A. Black, M. Li, and M. Tian, “Automatic Identification of Emotional Cues in Chinese Opera Singing,” in <em>13th Int. Conf. on Music </em><em>Perception and Cognition</em> (ICMPC-2014), 2014, pp. 250–255.</p> </blockquote> <p><strong>Annotations</strong></p> <p>The syllable boundary annotation is in Textgrid format (Praat). The annotation is done in both phrase-level and syllable-level. The syllable duration annotation is in cvs format. Please consult Readme text in both folders for further details. The parsing code of the annotation files is provided in ‘pycode’ folder. </p> <p><strong>Availability of the Dataset</strong></p> <p>The annotations and codes in this dataset are licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.</p> <p><strong>Contact</strong></p> <p>If you have any questions or comments about the dataset, please feel free to write to us.</p> <p>Rong Gong: rong<dot>gong<at>upf<dot>edu</p> <p>Rafael Caro Repetto: rafael<dot>caro<at>upf<dot>edu</p>
Jingju a cappella singing pitch contour segmentation ground truth dataset
<p>The dataset used in the paper:</p> <blockquote> <p>Gong, Rong; Yang, Yile; Serra, Xavier; Pitch Contour Segmentation for Computer-aided Jingju Singing Training Sound and Music Computing (SMC 2016), 2016, Hamburg, Germany</p> </blockquote> <p>is in "dataset" folder. The a cappella singing audio recordings are not contained in this folder due to their large size, please contact the paper authors to request them (rong.gong@upf.edu). In the "dataset" folder you can find:</p> <ol> <li>ground truth</li> <li>Jinging singing scores in .xml format used for estimating the bigram note transition probabilities.</li> </ol> <p>The ground truth annotation is used for:</p> <ul> <li>melodic transcription (male_12_pos_1 missing)</li> <li>parameter optimization,</li> <li>evaluating the StdCdLe thresholding and the overall segmentation performance.</li> </ul> <p>The subfolder "groundtruth" contains the following annotation for each jingju a cappella audio:</p> <ul> <li>file name: description (format)</li> <li>*_melodicTrans.csv: melodic transcription ground truth used for the evaluation (start_time pitch duration -).</li> <li>*_coarseSeg.csv: StdCdLe ground truth used for the parameter optimization and the evaluation (segmentation points).</li> <li>*_refinedSeg.csv: ground truth used for optimizing other parameters and the evaluation (start_time - duration).</li> <li>*_pitchtrack.csv: pitch track (contour) extracted by pYIN pitch-tracking algorithm (filename time pitch).</li> <li>*_monoNoteOut.csv: notes estimated by pYIN note-tracking algorithm (filename start_time duration pitch).</li> </ul> <p> </p> <p> </p>
University of Rochester Audio-Visual Solo Singing Performance (URSing) Dataset
<p>We introduce a dataset for facilitating audio-visual analysis of singing performances. The dataset comprises a number of songs where singers’ solo voices are recorded in isolation. For each song, we provide the high-quality audio recordings of the solo singing voice and mix with accompaniments, and the video recording of the upper body of the vocal soloist which contains facial expressions and lip movements. We anticipate that the dataset will be useful for developing audiovisual source separation systems. Note that some of the accompaniment tracks come with the backing vocals, which introduces extra challenges of developing an audio-based singing voice separation system, and encourages researchers to integrate the soloists’ visual information to facilitate the separation process. We also anticipate that the dataset will be useful for other multi-modal information retrieval techniques such as audiovisual expressions analysis, audio-visual correspondence, audiovisual lyrics transcription, etc.</p>
Annotated-VocalSet: A Singing Voice Dataset
<p>This dataset provides annotations for the <a href="https://doi.org/10.5281/zenodo.1442513">VocalSet dataset</a>, which is available online at</p> <pre><a href="https://doi.org/10.5281/zenodo.1442513">https://doi.org/10.5281/zenodo.1442513</a></pre> <p>.</p> <p>The annotations generated for the VocalSet audio files include fundamental frequency contour, note onset, note offset, the transition between notes, note F0, note duration, Midi pitch, and lyrics.</p> <p><a href="https://doi.org/10.5281/zenodo.1442513">VocalSet</a> consists of more than 10 hours of monophonic recorded audio of professional singers in a variety of vocal techniques (n = 17) and several singers (m = 20) with several WAV files (p = 3560). However, although several categories, including techniques, singers, tempo, and loudness, are considered in the dataset, the sung notes were not annotated. Therefore, this dataset aims to annotate VocalSet to make it a more powerful dataset for researchers.</p> <p>Details of the dataset are provided in the following academic journal paper.</p> <p><a href="https://www.mdpi.com/2076-3417/12/18/9257">Faghih, Behnam, and Joseph Timoney. 2022. "Annotated-VocalSet: A Singing Voice Dataset" <em>Applied Sciences</em> 12, no. 18: 9257. https://doi.org/10.3390/app12189257</a></p> <p>Please use the above paper to cite this dataset.</p>
Dataset: The Singing Machine Company, Inc. (MICS) Stock Performance
This dataset provides historical stock market performance data for specific companies. It enables users to analyze and understand the past trends and fluctuations in stock prices over time. This information can be utilized for various purposes such as investment analysis, financial research, and market trend forecasting.
Jingju a cappella singing dataset part2
<p>这个京剧清唱数据库包含有120个唱段、1265个唱句。此数据库是CompMusic项目所有数据库的一个组成部分(http://compmusic.upf.edu/corpora, http://compmusic.upf.edu/datasets)。CompMusic前期所使用的另一个京剧数据库可以在这里找到(https://doi.org/10.5281/zenodo.344932)。我们邀请了专业和业余的京剧演员参与到录音过程当中,大部分的京剧音乐元素都被囊括在了这个数据库中。此外,它还包含有每个唱段和每个唱句的元数据,以供自动演唱评价的研究使用。</p> <p>This is a jingju (also known as Beijing or Peking opera) a cappella singing audio dataset which consists of 120 arias, accounting for 1265 melodic lines. This dataset is also an extension our existing CompMusic jingju corpora (http://compmusic.upf.edu/corpora) and dataset (http://compmusic.upf.edu/datasets), for example, Jingju a cappella singing dataset part1 (https://doi.org/10.5281/zenodo.344932). Both professional and amateur singers were invited to the dataset recording sessions, and the most common jingju musical elements have been covered. This dataset is also accompanied by metadata per aria and melodic line annotated for automatic singing evaluation research purpose.</p> <p> </p> <p><strong>文件 Files:</strong></p> <ol> <li>wav.zip: audio files in .wav format, mono</li> <li>metadata.zip: aria and line level metadata</li> <li>annotation_version2.zip: line and syllable time boundaries and labels annotations, in Praat .textgrid format</li> <li>annotation_txt.zip: line and syllable time boundaries and labels annotations, in .txt format <ol> <li>*phrase_char: phrase-level time boundaries, labeled in Mandarin characters</li> <li>*phrase: phrase-level time boundaries, labeled in Mandarin pinyin</li> <li>*syllable: syllable-level time boundaries, labeled in Mandarin pinyin</li> </ol> </li> </ol> <p> </p> <p><strong>艺术家 Artists:</strong></p> <p>我们邀请了5位专业的京剧演员(中国戏曲学院,他们都有丰富的舞台表演和教学经验)和4位非艺术类高校京剧社团的业余京剧演员。</p> <p>We invited 5 professional singers from NACTA (National Academy of Chinese Theatre Arts, all of them have rich experience in stage performance and teaching) and another 4 amateur singers from jingju associations in non-art schools to the recording sessions. </p> <p> </p> <p><strong>伴奏 Accompaniment:</strong></p> <p>7位演员(3位专业和4位业余)跟随商业录音伴奏;另外2位专业演员由专业京胡乐手伴奏(中国戏曲学院)。</p> <p>7 singers (3 professional and 4 amateurs) were singing along with the accompaniment of commercial audio recordings; other 2 professional singers were accompanied by 2 professional <em>jinghu</em> players (NACTA).</p> <p><strong>数据库的覆盖性,完整性,质量和重复利用性 Coverage, completeness, </strong><strong>quality</strong><strong> and reusability:</strong></p> <ol> <li><em><strong>覆盖性: </strong></em>数据库包含三个主要的京剧行当 - 老生、旦和净;两个主要声腔 - 西皮和二黄,和一些附属声腔,比如四平调、南梆子;包含所有的有节拍的板式 - 原版、慢板、快板、二六、流水、三眼和他们的变化板式。<strong><em>Coverage</em></strong>: The dataset includes the three main role-types -<em> </em><em>laosheng</em><em>, dan</em> and<em> jing</em>; two main <em>shengqiang</em> - <em>xipi</em> and <em>erhuang</em>, and a few auxiliary ones, such as <em>sipingdiao</em><em>, </em><em>nanbangzi</em><em>;</em> the whole range of metered <em>banshi</em> - <em>yuanban</em><em>, </em><em>manban</em><em>, </em><em>kuaiban</em><em>, </em><em>erliu</em><em>, </em><em>liushui</em><em>, </em><em>sanyan</em> and its three variations.</li> <li><em><strong>完整性: </strong></em>数据库包含有录音和唱句层级的元数据,由Excel spreadsheet格式保存。对于录音层级,元数据包括唱段名、行当、声腔、板式、是否由京胡伴奏。对于唱句层级,每一句都包含行当、声腔、板式、上下句、唱词和所匹配的MusicXML曲谱(有需要曲谱请联系作者)。<strong><em>Completeness</em></strong>: The dataset contains the metadata of the recordings and annotations both at the recording and the line level, organized in separate spreadsheets. For the recordings, the metadata contains the title of the work in Chinese, role-type, <em>shengqiang</em><em>, </em><em>banshi</em>, whether it contains jinghu accompaniment. As for the lines, each of them is annotated with the role-type, <em>shengqiang</em><em>, </em><em>banshi</em><em>,</em> line type, that is, opening or closing, the lyrics for the whole line and the related score in the score collection (available on request).</li> <li><em><strong>质量: </strong></em>一小部分的录音带有中等程度的房间混响和轻微的背景噪声。其余的录音质量都很好。<strong><em>Quality</em></strong>: A small number of the recordings contain medium room reverberation and minor background noise. However, apart from those, the other recordings are dry, clean and of good quality.</li> <li><em><strong>重复利用性: </strong></em>所有数据库音频和元数据都由Creative Commons Attribution-NonCommercial 4.0 International方式授权。<strong><em>Reusability</em></strong>: All the audio and metadata files in this dataset are licensed under Creative Commons Attribution-NonCommercial 4.0 International.</li> </ol> <p> </p> <p><strong>标注 Annotation:</strong></p> <p>数据库包含一部分录音的唱句起始位置和音节起始位置标注,标注格式为Praat TextGrid。唱句标注包含有每一唱句的歌词,此歌词从曲谱提取,并不与实际演唱一致;音节标注包含拼音,经过作者修正,试图与演唱发音一致。标注的统计如下:</p> <ul> <li>老生唱句数量,音节数量,音节平均时长 (秒),音节时长标准差 (秒): 405, 3941, 1.32, 2.15</li> <li>旦唱句数量, 音节数量, 音节平均时长 (秒), 音节时长标准差 (秒): 467, 4394, 1.63, 3.25</li> <li>总体唱句数量, 音节数量, 音节平均时长 (秒), 音节时长标准差 (秒): 872, 8335, 1.48, 2.79</li> </ul> <p>The dataset contains the line and syllable boundary annotation for a part of recordings, in Praat TextGrid format. The line annotation contains the lyrics for each line, which is extracted from the score, and might not coherent with the actual singing; the syllable annotation contains pinyin, corrected by the author to be coherent with the actual singing. The statistics of the annotation are:</p> <ul> <li><em>laosheng</em> num. of lines, num. of syllables, average syllable duration (s), standard deviation (s): 405, 3941, 1.32, 2.15</li> <li><em>dan</em> num. of lines, num. of syllables, average syllable duration (s), standard deviation (s): 467, 4394, 1.63, 3.25</li> <li>Overall num. of lines, num. of lines, num. of syllables, average syllable duration (s), standard deviation (s): 872, 8335, 1.48, 2.79</li> </ul> <p> </p> <p><strong>引用 Citation:</strong></p> <p>如需更多信息,请参考下面论文;如果您在工作中使用该数据库,请引用下面论文:</p> <p>For more information, please refer the following publication and If you use this dataset in your work, please cite the following publication:</p> <blockquote> <p>Rong Gong, Rafael Caro Repetto, Xavier Serra, “Creating an A Cappella Singing Audio Dataset for Automatic Jingju Singing Evaluation Research,” in 4th International Digital Libraries for Musicology workshop (DLfM 2017), Shanghai, China.</p> </blockquote> <p> </p> <p><strong>协议 License:</strong></p> <p><a href="https://creativecommons.org/licenses/by-nc/4.0/">Creative Commons Attribution-NonCommercial 4.0</a></p> <p> </p> <p><strong>联系方式 Contact information</strong>:</p> <p><em>如果任何问题,请联系作者 </em><em>If you have any question, please contact the authors:</em></p> <p>龚嵘 Rong Gong: Email - rong<dot>gong<at>upf<dot>edu, Wechat id - gongr86</p> <p>贵云飞 Rafael Caro Repetto: Email - rafael<dot>caro<at>upf<dot>edu </p> <p> </p> <p><em>如果您想联系京剧演员 If you want to contact the jingju</em><em> singers:</em></p> <p>廖佳尼 Jiani Liao: Wechat id - v1307624197</p> <p>邵雅昆 Yakun Shao: Wechat id - S_yakun-</p> <p> </p> <p><em>或京胡乐手 Or jinghu</em><em> players:</em></p> <p>张蓝天 Lantian Zhang: Wechat id - tian576632395</p>
VocalSet: A Singing Voice Dataset
<p><strong>NEW IN VocalSet 1.2:</strong> We now have 3 file organization versions:</p> <ol> <li>Files organized by singer</li> <li>Files organized by technique</li> <li>Files organized by vowel</li> </ol> <p>We hope that this will ease the process of training and testing models using these different attributes of the dataset.</p> <p> </p> <p><strong>Overview:</strong></p> <p>We present VocalSet, a singing voice dataset consisting of 10.1 hours of monophonic recorded audio of professional singers demonstrating both standard and extended vocal techniques on all 5 vowels. Existing singing voice datasets aim to capture a focused subset of singing voice characteristics, and generally consist of just a few singers. VocalSet contains recordings from 20 different singers (9 male, 11 female) and a range of voice types. VocalSet aims to improve the state of existing singing voice datasets and singing voice research by capturing not only a range of vowels, but also a diverse set of voices on many different vocal techniques, sung in contexts of scales, arpeggios, long tones, and excerpts.</p> <p>We have included two .txt files 'train_singers_technique.txt 'and 'test_singers_technique.txt' in which you will find a list of the singers we used to train and test our technique classifier on. 'DataSetVocalises.pdf' contains the sheet singers sang from in their recording sessions. 'readme-anon.txt' contains more information about the dataset, including the mapping from filename to singer voice type as well as more information on the vocalises that will help you map files to sheet music. Enjoy and please cite accordingly!</p>
MedleyVox: An Evaluation Dataset for Multiple Singing Voices Separation
<p>MedleyVox is an evaluation dataset for multiple singing voices separation, which consists of 381 segments (1.1hour), containing 23 songs from MedleyDB v1 and v2 (https://medleydb.weebly.com).</p> <p> </p> <p>MedleyVox contains 1) unison, 2) duet, 3) main vs. rest (folder name 'rest') , and 4) N-singing ('unison' + 'duet' + 'rest', please check our dataloader code for N-singing in https://github.com/jeonchangbin49/MedleyVox/blob/main/svs/data/test_dataset.py) categories explained in our ICASSP 2023 paper. For more details, please check our paper (https://arxiv.org/pdf/2211.07302.pdf) and code repository (https://github.com/jeonchangbin49/MedleyVox).</p> <p> </p> <p>Acknowledgment</p> <p>We are grateful to Rachel Bittner, the author of the original MedleyDB data, for allowing us to publish the MedleyVox dataset.</p> <p> </p> <p>License</p> <p>This work is licensed under a <a href="https://creativecommons.org/licenses/by-nc-sa/4.0/">Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License</a>.</p>
Underlying REVISED dataset for the study: Singing and music making: Physiological responses across early to later stages of dementia
<p>These files contain the REVIDED underlying data for the study, Singing and Music Making: Physiological Responses Across Early to Later Stages of Dementia; physiological data from Study 1 and Study 2, and video recording engagement scores from Study 2.</p> <p><strong>in Study 2, the electrodermal readings (EDA) for the Slow Music readings in session 1 had become corrupted. The correct EDA data for this session was uploaded on 31.3.24.</strong></p> <p>The project was funded by the Wellcome Trust as part of The Hub Award at the Wellcome Collection, London. The article that these data are based on can be here: https://wellcomeopenresearch.org/articles/6-150/v1, Wellcome Open Research, 6:150.</p> <p>The abstract from the accompanying article:</p> <p><strong>Background</strong>: Music based interventions have been found to improve the wellbeing of people living with dementia. Research to date has primarily used psychometric questionnaires and qualitative interviews to determine impact and efficacy. More recently there has been an interest in exploring if psychophysiological measures could provide additional information about how music, singing and other arts activities impact this population. Physiological responses can provide additional evidence about an individual’s experience of an activity and may be particularly useful for people who are experiencing difficulties with communication. </p> <p><strong>Methods:</strong> This multiple-case study design drew on previously collected, unanalysed archival data and explored the physiological responses of nine people with mild-to-moderate dementia during a singing group, and six people in the later stages of dementia during an interactive music group. Medical grade Empatica E4 ™ wearable wristbands provided information on heart rate (HR), electrodermal activity (EDA), movement (ACC) and skin temperature (ST). The interactive music group was video recorded using non-intrusive Fly 360-degree cameras™ in order to provide additional group interactive information about engagement. </p> <p><strong>Results</strong>: Physiological responses were analysed using simulation modelling analysis (SMA) within individual case studies. Participants in the singing group showed an increase in EDA and HR as the session began. HR and ST increased during faster paced songs. EDA, movement and engagement were all higher during an interactive music group than during a control session (music listening). EDA and ST increased and in contrast to the responses during singing, HR decreased as the sessions began. EDA was higher during slower music, however this was less consistent in the more interactive intervention sessions than the control. There were no consistent changes in HR and movement responses during different styles of music. Physiological responses peaked during familiar music, interactions, physical touch in addition to times that participants appeared disengaged. </p> <p><strong>Conclusion</strong>: Non-intrusive physiological measures obtained from easily worn wristband devices may provide valuable information about the experiences of people living with dementia who participate in arts and other activities, particularly for those in later stages when verbal communication may be more difficult and it is no longer possible to complete psychometric questionnaires. However, whenever feasible, they should be used in conjunction with other measures to develop a more nuanced understanding of these experiences. Future research should consider using physiological measures with video-analysis and observational measures, where possible, to explore further how engagement in specific activities, wellbeing and physiology interact. This may provide valuable information for further development of activities and services for those living with dementias across different levels of impairment.</p>
Dataset for Interspeech 2018 submission: Singing voice phoneme segmentation by hierarchically inferring syllable and phoneme onset positions
<p>This dataset contains the materials for training, testing the joint and HSMM models mentioned in the paper "<em>Singing voice phoneme segmentation by hierarchically inferring syllable and phoneme onset positions"</em>.</p> <p>The filename list of this dataset can be found in the function <em>get_train_test_recordings_joint()</em> of <em>./general/trainTestSeparation.py</em> file. The dataset contains the Praat TextGrids and .wavs of the variables: <em>train_primary_school, val_primary_school</em> and <em>test_primary_school</em>. For accessing other datasets such as <em>train_nacta_2017, train_nacta</em> and <em>train_sepa</em>, please download them from the links:</p> <p>jingju dataset part1: <a href="https://zenodo.org/record/1185154">https://zenodo.org/record/1185154</a></p> <p>jingju dataset part2: <a href="https://doi.org/10.5281/zenodo.842229">https://doi.org/10.5281/zenodo.842229</a></p> <p>Once you have downloaded these three datasets, you need to set the paths in <em>./general/filePathShared.py</em>.</p> <p>Set <em>path_jingju_dataset</em> to the parent path of these three datasets.</p> <p>Set <em>primarySchool_dataset_root_path</em> to the path of the interspeech2018 dataset (the current dataset).</p> <p>Set <em>nacta_dataset_root_path</em> to the path of the jingju dataset part1.</p> <p>Set <em>nacta2017_dataset_root_path</em> to the path the jingju dataset part2.</p> <p>For more information on this paper, please refer to the Github page: <a href="https://github.com/ronggong/interspeech2018_submission01">https://github.com/ronggong/interspeech2018_submission01</a></p> <p> </p>
Jingju a cappella singing dataset part3
<p>这是京剧清唱数据库的第三部分。这个部分集中于音乐教育的应用。对每个唱段,我们采集了老师和学生的录音。</p> <p>This is the 3rd part of the jingju a cappella singing dataset. This part focus on the music education application. For this purpose, we collected both "teacher" and "students" recordings for each aria.</p> <p><strong>文件 Files:</strong></p> <ol> <li>wav_left.zip: audio files in .wav format, mono</li> <li>textgrid.zip: line, syllable and phoneme time boundaries and labels, in Praat .textgrid format</li> <li>annotation_txt.zip: line, syllable and phoneme time boundaries and labels, in .txt format <ol> <li>*phrase_char: phrase-level time boundaries, labeled in Mandarin characters</li> <li>*phrase: phrase-level time boundaries, labeled in Mandarin pinyin</li> <li>*syllable: syllable-level time boundaries, labeled in Mandarin pinyin</li> <li>*phoneme: phoneme-level time boundaries, labeled in X-SAMPA</li> </ol> </li> <li>arias.ods: spreadsheet containing the detailed information of each recording -- role-type, shengqiang, banshi, singer and melodic line, syllable, phoneme numbers.</li> </ol> <p> </p> <p><strong>艺术家 Artists:</strong></p> <p>示范录音取自3位年轻的专业京剧演员。学生录音取自几位非专业小学生和非艺术类大学的学生。</p> <p>Teacher recordings are recorded by 3 professional young jingju performers. Student recordings are performed by several non-jingju professional primary school students and some non-art university students. </p> <p> </p> <p><strong>标注 Annotation:</strong></p> <p>数据库包含一部分录音的唱句起始位置和音节起始位置标注,标注格式为Praat TextGrid。唱句标注包含有每一唱句的歌词,此歌词从曲谱提取,并不与实际演唱一致;音节标注包含拼音,音素标注为X-SAMPA,经过作者修正,试图与演唱发音一致。</p> <p>The dataset contains the line and syllable boundary annotation for a part of recordings, in Praat TextGrid format. The line annotation contains the lyrics for each line, which is extracted from the score, and might not coherent with the actual singing; the syllable annotation contains pinyin, phoneme annotation uses X-SAMPA, corrected by the author to be coherent with the actual singing.</p> <p>标注X-SAMPA格式和其他信息可以参考以下链接:</p> <p>Annotation format, units, parsing code and other information please refer to:</p> <p><a href="https://github.com/MTG/jingjuPhonemeAnnotation">https://github.com/MTG/jingjuPhonemeAnnotation</a></p> <p> </p> <p><strong>协议 License:</strong></p> <p><a href="https://creativecommons.org/licenses/by-nc/4.0/">Creative Commons Attribution-NonCommercial 4.0</a></p> <p> </p> <p><strong>联系方式 Contact information</strong>:</p> <p><em>如果任何问题,请联系作者 </em><em>If you have any question, please contact the author:</em></p> <p>龚嵘 Rong Gong: Email - rong<dot>gong<at>upf<dot>edu, Wechat id - gongr86</p>
Jingju a cappella singing voice test dataset for "An efficient deep learning model for musical onset detection"
<p>Jingju a cappella singing voice test dataset used in the paper "An efficient deep learning model for musical onset detection".</p> <p>Arxiv paper link: <a href="https://arxiv.org/abs/1806.06773">https://arxiv.org/abs/1806.06773</a></p> <p>Supplementary information and code for the paper: <a href="https://github.com/ronggong/musical-onset-efficient">https://github.com/ronggong/musical-onset-efficient</a></p> <p><strong>Content:</strong></p> <ol> <li>ismir_2018_dataset_for_reviewing.zip: audio, syllable boundary and label annotation</li> <li>jingju dataset train test split filenames.xlsx: train and test split filename list</li> </ol> <p><strong>Citation:</strong></p> <pre>@article{gong2018towards, title={Towards an efficient deep learning model for musical onset detection}, author={Gong, Rong and Serra, Xavier}, journal={arXiv preprint arXiv:1806.06773}, year={2018} } </pre> <p><strong>Contact:</strong></p> <p>Rong Gong: rong.gong<at>upf.edu</p>
Choral Singing Dataset
<p><strong>Choral Singing Dataset</strong></p> <p>This dataset was presented at the 15th ICMPC/10th ESCOM conference together with the paper: </p> <p>Cuesta, H., Gómez, E., Martorell, A., Loáiciga, F. (2018). <em>Analysis of Intonation in Unison Choir Singing</em>. In <em>Proceedings of the 15th International Conference on Music Perception and Cognition.</em></p> <p>It contains the individual audio recordings of 16 singers of the Anton Bruckner Choir from Barcelona (Spain) performing 3 different pieces <em>a cappella, </em>together with their associated MIDI files.</p> <p>Singers were recorded in groups of four (4 singers per choir section), with individual close microphones with cardioid polar pattern for directivity purposes. The performance was virtually conducted by the conductor of the choir through a video, which was displayed in all the recording sessions for synchronization purposes. Singers also had the possibility to hear a piano reference (through headphones) for tuning purposes.</p> <p>Three pieces were selected based on the choir repertoire and on the specific needs of the study, which were basically related to the language of the lyrics. These are the pieces we chose:</p> <ul> <li><em>Locus Iste</em>, written by Anton Bruckner (Latin).</li> <li><em>Niño Dios</em>, written by Francisco Guerrero (Spanish).</li> <li><em>El Rossinyol</em>, popular Catalan song.</li> </ul> <p>Overall, this dataset contains, for each of these pieces, the tracks of each individual singer (16 singers), together with frame-wise f0 annotations, and note annotations and semi-synchronized MIDI files for each choir section.</p> <p>Having the individual tracks also allows researchers to create the unison mix for each section, as well as the whole choir performance.</p> <p>The dataset covers the frequency range between 87 Hz and 783 Hz and is especially dense between 150 and 450 Hz. The notes have durations that range between 0.15 and 6.21 seconds, with an average of 0.84 seconds. </p> <p>If using this data in your research, please cite the aforementioned paper, which is also available for download. For any question or comment, please contact the first author.</p> <p>The dataset is compressed into a zip file. In the decompressed folder there is a README file with all the information regarding the annotation files and the filenames.</p>
Dataset for: Vermilion flycatchers avoid singing during sudden peaks of anthropogenic noise
<p>Dataset from a playback experiment with free-living vermilion flycatchers where we show that they stop singing when traffic noise levels suddently increase (as would occur when a car passes by). This behavior may be a strategy to increase the probability of detection by singing in periods when noise levels are lower. Males' responses toward the noise playbacks were also related with the level of artificial light at night (ALAN) in males' territories, where males living in brighter territories shortened their song bout length to a lesser extent during noise exposure than those males living in territories with lower levels of ALAN, suggesting ALAN may decrease the extent at which traffic noise affects male singing behavior.</p>
ChoralSynth: Synthetic Dataset of Choral Singing
<p><strong>Overview</strong></p><p>ChoralSynth is a synthesized dataset of 20 multitrack choral songs curated by carefully listening and analyzing a set of synthetic choral songs generated using one of the state-of-the-art synthesizers. The resulting dataset can serve as a valuable resource for various MIR research endeavors like source separation, melodic analysis, chord analysis, rhythmic analysis amongst others.</p>
High-quality Image (NIR and RGB) Dataset Synchronized With Contact Vital Sings Recordings and Clinical Data of Stratified Healthy Population. Algorithms and AI Models to Obtain a Set of Vital Signs Im
ClinicalTrials.gov study NCT05947721. IPD Sharing: NO. Countries: 1. Publications: 2.
singing voice detection dataset ALL-Pub-SVD-In-One
<p>singing voice detection datasets:</p> <p>it contains the public dataset of Jamendo Dataset, and the RWC dataset , and the iKala dataset, and the MIR-1K dataset , and teh MedlyDB dataset.</p> <p>Jamendo are same with the raw.</p> <p>RWC use the simply mode same with the lable in Jamendo just sing or nosing.</p> <p>iKala and MIR-1K are full wavfiles with the generated lables from the PitchLables. 0 for nosing and other for sing.</p> <p>MedlyDB use part wavfiles from the RAW dataset and also generated lables from the MELODY_2 lables with 0 for nosing and other for nosing.</p> <p> </p>
Singing Voice Audio Dataset
<p>This dataset is for the purpose of the analysis of singing voice. It is our hope that the publication of this dataset will encourage further work into the area of singing voice audio analysis by removing one of the main impediments in this research area - the lack of data (unaccompanied singing).</p> <p>It contains over 70 original vocal recordings by 28 professional, semi-professional and amateur singers. Singing style is predominantly Chinese Opera but some recordings are Western Opera. All recordings are 44.1 KHz sample rate and have been amplitude normalised.</p>
ScienceDex guides
Understand access before you commit
These curated guides explain access requirements, typical timelines, costs, and reuse considerations for widely used research datasets.
Allen Brain Atlas
Allen Brain Atlas is an Allen Institute collection of brain map atlases, datasets, APIs, and analysis tools covering mouse, human, and non-human primate brain resources.
Annotated Behaviour and Observability Dataset (ABODe)
ABODe is a University of Edinburgh DataShare dataset for behavior classification in group-housed mice using home-cage video, identities, bounding boxes, ground-plate positions, and annotator labels.
DANDI Archive for NWB datasets
DANDI is a BRAIN Initiative archive for publishing and sharing neurophysiology data, including electrophysiology, optophysiology, and behavioral data packaged as NWB and related standards.
International Brain Laboratory public data
The International Brain Laboratory public data releases expose standardized mouse decision-making experiments, including Neuropixels recordings, widefield calcium imaging, behavior, and session metadata accessed through the ONE API.
OpenNeuro
OpenNeuro is a free, open platform for sharing neuroimaging datasets, with public search, dataset pages, and download paths for web, S3, DataLad, and the OpenNeuro CLI.