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.
106
datasets available to search
ShareScore release 0.9.0
Dataset results
106 results for “Audio Dataset”
IDMT-SMT-Audio-Effects Dataset
<p>The IDMT-SMT-Audio-Effects database is a large database for automatic detection of audio effects in recordings of electric guitar and bass and related signal processing.</p> <p>The overall duration of the audio material is approx. 30 hours.</p> <p>The dataset consists of 55044 WAV files (44.1 kHz, 16bit, mono) with single recorded notes:</p> <ul> <li>20592 monophonic bass notes</li> <li>20592 monophonic guitar notes</li> <li>13860 polyphonic guitar sounds</li> </ul> <p>Overall, 11 different audio effects are incorporated:</p> <ul> <li>feedback delay</li> <li>slapback delay</li> <li>reverb</li> <li>chorus</li> <li>flanger</li> <li>phaser</li> <li>tremolo</li> <li>vibrato</li> <li>distortion</li> <li>overdrive</li> <li>no effect (unprocessed notes/sounds)</li> </ul> <p>2 different electric guitars and 2 different electric bass guitars, each with two different pick-up settings and up to three different plucking styles (finger plucked - hard, finger plucked - soft, picked) were used for recording.</p> <p>The notes cover the common pitch range of a 4-string bass guitar from E1 (41.2 Hz) to G3 (196.0 Hz) or the common pitch range of a 6-string electric guitar from E2 (82.4 Hz) to E5 (659.3 Hz).</p> <p>Effects processing was performed using a digital audio workstation and a variety of mostly freely available effect plugins.</p> <p>To organize the database, lists in XML format are used, which record all relevant information and are provided with the database as well as a summary of the used effect plugins and parameter settings.</p> <p>In addition, most of this information is also encoded in the first part of the file name of the audio files using a simple alpha-numeric encoding scheme. The second part of the file name contains unique identification numbers. This provides an option for fast and flexible structuring of the data for various purposes.</p>
MARVEL - Malta Audio Visual Anomaly Dataset (MAVAD)
<p>The raw audio-video data was collected from two locations on the island of Malta one in Zejtun, a town close to the industrial region on the eastern coast, and another in Mgarr, a rural town on the western coast. Three AV cameras were deployed, two in Zejtun (Zejtun Scrapyard and Zejtun Field) and one in Mgarr, resulting in three datasets with the same titles.</p> <p> </p>
CFAD: A Chinese Dataset for Fake Audio Detection
<p>Fake audio detection is a growing concern and some relevant datasets have been designed for research. However, there is no standard public Chinese dataset under complex conditions.</p> <p>In this paper, we aim to fill in the gap and design a Chinese fake audio detection dataset (CFAD) for studying more generalized detection methods. Twelve mainstream speech-generation techniques are used to generate fake audio. To simulate the real-life scenarios, three noise datasets are selected for noise adding at five different signal-to-noise ratios, and six codecs are considered for audio transcoding. CFAD dataset can be used not only for fake audio detection but also for detecting the algorithms of fake utterances for audio forensics. Baseline results are presented with analysis. The results that show fake audio detection methods with generalization remain challenging. The CFAD dataset is publicly available on GitHub https://github.com/ADDchallenge/CFAD</p> <p> </p> <p>CFAD dataset considers 12 types of fake audio, 11 of which are generated by different speech synthesis techniques and the remaining one is partially fake type. Partially fake audio is completely different from synthesis speech and thus can better evaluate the generalization of the detection model to unknown types. The real audio is collected from 6 different corpora to increase the diversity of real category distributions, which makes model less prone to artifact from a single database. For robustness evaluation, we additionally simulate background noise and media codecs that might occur in real life and provide detailed labels, including fake type, real source, noise type, signal noise ratio (SNR), and media codecs. Overall, CFAD dataset consists of three different versions, named clean, noisy, and codec versions.</p> <p>Each version of the dataset is divided into disjoint training, development, and test sets in the same way. There is no speaker overlap across these three subsets. Each test set is further divided into seen and unseen test sets. Unseen test sets can evaluate the generalization of the methods to unknown types. It is worth mentioning that both real audio and fake audio in the unseen test set are unknown to the model.</p> <p>For the noisy speech part, we select three noise databases for simulation. Additive noises are added to each audio in the clean dataset at 5 different SNRs. The additive noises of the unseen test set and the remaining subsets come from different noise databases.</p> <p>For the codec speech part, we select six different codecs. Two of them are applied for unseen test set.</p> <p>In each version (clean, noisy, and codec versions) of the CFAD dataset, there are 138400 utterances in training set, 14400 utterances in development set, 42000 utterances in seen test set, and 21000 utterances in unseen test set.</p> <p> </p> <p>Clean Real Audios Collection</p> <p>From the point of eliminating the interference of irrelevant factors, we collect clean real audios from two aspects: 5 open resources from OpenSLR platform (http://www.openslr.org/12/) and one self-recording dataset. </p> <p> </p> <p>Clean Fake Audios Generation</p> <p>We select 11 representative speech synthesis methods to generate the fake audios and one partially fake audios.</p> <p> </p> <p>Noisy Audios Simulation</p> <p>Noisy audios aim to quantify the robustness of the methods under noisy conditions. To simulate the real-life scenarios, we artificially sample the noise signals and add them to clean audios at 5 different SNRs, which are 0dB, 5dB, 10dB, 15dB and 20dB. Additive noises are selected from three noise databases: PNL 100 Nonspeech Sounds, NOISEX-92, and TAU Urban Acoustic Scenes.</p> <p> </p> <p>Audio Transcoding</p> <p>The Codec version aims to quantify the robustness of the methods under different format conversions. We select a total of six codecs. For the training, development, and seen test sets in codec version, mp3, flac, ogg, and m4a are used. For the unseen test set of the codec version, aac, and wma are used.</p> <p>Audio transcoding operation is operated on the audio in the clean version. Each clean audio will be randomly transformed with one of the candidate codecs and converted back to original WAV files using ffmpeg toolkits.</p> <p> </p> <p> </p> <p>This data set is licensed with a CC BY-NC-ND 4.0 license.</p> <p>You can cite the data using the following BibTeX entry.</p>
Interruption Audio & Transcript: Derived from Group Affect and Performance Dataset
<p><strong>Licensing</strong></p> <p>This dataset is adapted from the Group Affect and Performance dataset which is released under the Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) license. <a href="https://creativecommons.org/licenses/by-nc/4.0/">https://creativecommons.org/licenses/by-nc/4.0/</a></p> <p> </p> <p><strong>Description</strong></p> <p>This dataset contains the audio files containing manually annotated cases of overlapped utterances, classified into True Interruptions and False Interruptions. It is derived from the <a href="https://github.com/gmfraser/gap-corpus/tree/master">Group Affect and Performance</a> dataset created by the University of the Fraser Valley, Canada. Original conversation transcripts and audio files have been supplied for context. The Group Affect and Performance dataset provides a rich source of interruptions and overlapped utterances in general, yielding 200 True Interruptions from 355 instances of overlapped utterances in the 14 Group meetings which were annotated.</p> <p> </p> <p><strong>Structure</strong></p> <p>This dataset is structured into three parts:</p> <p>1. data.json contains a list of all instances of overlapped utterances, classified into ‘interruption’ and ‘non-interruption’ corresponding to True and False Interruptions respectively. Each instance is uniquely identified by the Group in which it occurred, the speaker and the starting time of the utterance.</p> <p>2. The 'audio' directory contains the audio of each instance of overlapped utterances corresponding to those found in data.json. The naming convention of the files is as such: ‘Group [group number]: [utterance start time] - [utterance end time].wav’.</p> <p>3. Also included is a copy of the original dataset which includes the full audio and transcript. This allows the full meeting to be heard and any context for interruptions to be evaluated.</p> <p>Note that directories 2. and 3. can be accessed by unzipping audio-and-transcripts.zip.</p> <p> </p> <p><strong>Data Collection Protocol</strong></p> <p>Of paramount importance to our process are the definitions of an overlapped utterance and a True Interruption. A False Interruption is simply an overlapped utterance which is not a True Interruption. These definitions directly impact the dataset; for overlapped utterance it informs which data points are included in our dataset and for True Interruption it informs the classes assigned to each sample.</p> <p>In defining an overlapped utterance, our primary aim is to create an overarching class encompassing interruptions and all instances that could be deemed a True Interruption. For this reason, we omit cases where the timing misplaced speech and early-onset responses.</p> <p>An overlapped utterance is defined as an instance where one interlocutor provides speech or noise during another interlocutor’s speech, creating an overlap that may be deemed a possible interruption when considering its timing alone. For this reason we omit cases of where the timing indicates misplaced speech or early-onset responses.</p> <p>Our definition of True Interruption is an instance where an interrupting party intentionally attempts to take over a turn of the conversation from an interruptee and, in doing so, creates an overlap in speech.</p> <p>As previously mentioned, due to the ‘intent’ part of this definition, we avoid cases of misplaced speech and early-onset responses. The former is enforced by not considering cases of overlapped speech which begin within 300ms of each other since this is an estimate for the average human reaction time of articulating a vowel in response to a speech stimuli. The latter is enforced by not considering speech starting within the last 10% of first utterance in the overlapped speech. Note that this approach fails to filter out all cases of misplaced speech, so we manually remove the remaining instances.</p> <p> </p> <p><strong>Methodology</strong></p> <p>Three main steps were taken to produce this dataset:</p> <p>1. Parsing the transcripts for cases of overlapping speech</p> <p>2. Manually annotating these cases per our protocol and adding them to data.json</p> <p>3. Extracting audio samples from data.json and adding them to the audio folder</p> <p> </p> <p>If you use this dataset, please cite the following paper:</p> <p> </p> <p>Doyle, D.; Şerban, O. Interruption Audio & Transcript: Derived from Group Affect and Performance Dataset. <em>Data</em> <strong>2024</strong>, <em>9</em>, 104. https://doi.org/10.3390/data9090104</p> <p> </p> <p>@article{data9090104,</p> <p>AUTHOR = {Doyle, Daniel and Şerban, Ovidiu},</p> <p>TITLE = {Interruption Audio & Transcript: Derived from Group Affect and Performance Dataset},</p> <p>JOURNAL = {Data},</p> <p>VOLUME = {9},</p> <p>YEAR = {2024},</p> <p>NUMBER = {9},</p> <p>ARTICLE-NUMBER = {104},</p> <p>URL = {https://www.mdpi.com/2306-5729/9/9/104},</p> <p>ISSN = {2306-5729},</p> <p>DOI = {10.3390/data9090104}</p> <p>}</p>
Text to audio grounding (TAG) dataset: AudioGrounding
<p>AudioGrounding dataset, including audio files and timestamp annotations.</p><p>Changes in version 2: The train/validation/test sets are re-split. The validation and test annotations are refined.</p><p> </p><p>----------------------------------------------------------</p><p><strong>References</strong></p><p>[1] Xuenan Xu, Heinrich Dinkel, Mengyue Wu and Kai Yu. "Text-to-audio grounding: Building correspondence between captions and sound events." In <i>Proceedings of IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)</i>. IEEE, 2021, pp. 606-610.</p><p>[2] Xuenan Xu, Mengyue Wu, and Kai Yu. "Investigating Pooling Strategies and Loss Functions for Weakly-Supervised Text-to-Audio Grounding via Contrastive Learning." In <i>Proceedings of IEEE International Conference on Acoustics, Speech, and Signal Processing Workshops (ICASSPW)</i>. IEEE, 2023, pp. 1-5.</p>
Accompanying dataset for "Nappe oscillations on free-overfall structures, data from laboratory experiments (audio and video)"
<p>This dataset accompanies the manuscript "Nappe Oscillations on Free-Overfall Structures: Data from Laboratory Experiments" submitted to Scientific Data.</p> <p>This dataset contains raw audio and video data, which complement the dataset uploaded at: <a href="https://zenodo.org/record/3381078#.XlUGVSFKiUk">https://zenodo.org/record/3381078#.XlUGVSFKiUk</a></p> <p>The names of the folders describe each one of the 52 experiments, with respect to the submitted paper in Scientific Data:</p> <p>- M1 and M2 denote Model 1 and Model 2, respectively.</p> <p>- C and UC denote confined and unconfined nappe, respectively.</p> <p>- QR, THR, HR, R, and RR denote the crest type of the weir as explained in the paper.</p> <p>- W is the width of the crest and L is the falling height.</p>
EmoSynth: The Emotional Synthetic Audio Dataset
<p>The ability of sound to enhance human wellbeing has been known since ancient civilisations, and methods can be found today across domains of health and within a variety of cultures. </p> <p>EmoSynth is a dataset of 144 audio files which have been labelled by 40 listeners for their the perceived emotion, in regards to the dimensions of Valence and Arousal.</p> <p>Results on the dataset show that Arousal does correlate moderately to fundamental frequency, and that the sine waveform is perceived as significantly different to square and sawtooth waveforms when evaluating perceived Arousal. The general results suggest that isolated synthetic audio can be modelled as a means of evoking affective states of emotion.</p> <p>When using this version of the data for a publication, please cite the following study: </p> <pre>@inproceedings{Baird:2018:PEI:3243274.3243277, author = {Baird, Alice and Parada-Cabaleiro, Emilia and Fraser, Cameron and Hantke, Simone and Schuller, Bj\"{o}rn}, title = {The Perceived Emotion of Isolated Synthetic Audio: The EmoSynth Dataset and Results}, booktitle = {Proceedings of the Audio Mostly 2018 on Sound in Immersion and Emotion}, series = {AM'18}, year = {2018}, isbn = {978-1-4503-6609-0}, location = {Wrexham, United Kingdom}, pages = {7:1--7:8}, articleno = {7}, numpages = {8}, url = {http://doi.acm.org/10.1145/3243274.3243277}, doi = {10.1145/3243274.3243277}, acmid = {3243277}, publisher = {ACM}, address = {New York, NY, USA}, keywords = {affective computing, perception, sound healing, synthetic audio, wellbeing}, } </pre> <p> </p>
A dataset for high-level activity recognition based on low level audio events
<p>The high level activities are:<br> - kitchencleanup<br> - music<br> - no activity<br> - other activity<br> - talk<br> - tv</p> <p>Each recording of low-level audio events is stored in a separate file.</p> <p>Files are organized in 6 folders, each folder corresponding to a separate file.</p> <p>The format of is file is json-like. In particular, each row has the following format:</p> <p>{"prob": 0.88557562121157585, "energy": 0.024511212402412885, "t": 1485110417, "event": "speech"}</p> <p>This dataset can be evaluated with the python code metaClassifier/evaluate.py of the AUOR repository:<br> https://github.com/tyiannak/AUROS</p>
Dataset for honey bee audio detection
<p>The database contains recordings of bees and drones in wav format, without compression. The sampling frequency is 44.1kHz.</p><p>The database contains 10000 one-second recordings of bees and 1700 one-second recordings of drones.</p>
Audio Caption Dataset (Hospital & Car)
<p>This dataset consists of the Hospital scene of our Audio Caption dataset. Details can be seen in our paper <a href="https://arxiv.org/abs/1902.09254">Audio Caption: Listen and Tell</a> published at ICASSP2019. </p> <p>Car scene, detailed in <a href="https://arxiv.org/abs/1905.13448">Audio Caption in a Car Setting with a Sentence-Level Loss</a> published at ISCSLP 2021.</p> <p>Original captions in Mandarin Chinese, with English translations provided. </p>
Dízi Fingering Dataset: Acoustic Impedance and Audio Spectra, with Sound Files
<p>This database presents acoustic impedance spectra measured on a G-key Dízi (Chinese transverse flute): a total of 35 fingerings (both standard and alternative fingerings) for 24 semi-chromatic notes ranging from D5 to G7 (i.e. the instrument’s typical playing range), for both with the membrane attached and without. Accompanying each fingering entry is the sound recording and corresponding audio spectra.</p> <p> </p> <p>An interactive graphical version of this database is also presented at <a href="https://acoustics.sutd.edu.sg/dizi-impedance/">https://acoustics.sutd.edu.sg/dizi-impedance/</a> </p> <p> </p> <p>The impedance measurements (using the “<a href="https://apc01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fnewt.phys.unsw.edu.au%2Fjw%2Freprints%2FDickensSmithWolfe.pdf&data=04%7C01%7Cyixian_tan%40sutd.edu.sg%7C7763bf2bc6eb4f9c849908da09080e5d%7C3476b776e9904f72b95062489831623d%7C0%7C0%7C637832227461557470%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&sdata=AfSXOQEqxZxbz%2BPJ7ZnbtAuSZn6xF1YBBUmQenjRZmA%3D&reserved=0">three-microphone two-calibration</a>” technique) were conducted at the Acoustics Lab, School of Physics, UNSW, Sydney Australia. The authors are grateful to John Smith and Joe Wolfe for hosting us. The acoustic impedance measurement hardware and software was developed by Paul Dickens and we used a version of the software subsequently modified by Noel Hanna.</p>
HEAR NeurIPS 2021 Datasets (Holistic Evaluation of Audio Representations)
<p><strong>NOTES:</strong></p> <p><strong>* On Zenodo, please make sure you download datasets version 2021.3, not earlier versions. (2021.3 added the Vocal Imitations dataset at 48KHz. 2021.2 updated the tfds datasets.)</strong></p> <p><strong>* The datasets have different open licenses. Please see LICENSE.txt for each individual dataset's license.</strong></p> <p>These are the evaluation tasks for the HEAR (Holistic Evaluation of Audio Representations) 2021 NeurIPS challenge.</p> <p>The aim of this challenge is to develop a general-purpose audio representation that provides a strong basis for learning in a wide variety of tasks and scenarios. The HEAR 2021 challenge invites you to create an audio embedding that is as holistic as the human ear, i.e., one that performs well across a variety of everyday domains: What approach best generalizes to a wide range of downstream audio tasks without fine-tuning?<br> <br> HEAR 2021 evaluates audio representations using a benchmark suite across a variety of domains, including speech, environmental sound, and music.</p> <p>For more information, see the <a href="https://neuralaudio.ai/hear2021-holistic-evaluation-of-audio-representations.html">HEAR 2021 website</a> and upcoming PMLR journal article.</p> <p>Datasets were all normalized to a common human-readable format using <a href="https://github.com/neuralaudio/hear-preprocess">hearpreprocess</a>. Until 2022-04-01, datasets will be mirrored at <a href="https://data.neuralaudio.ai/">data.neuralaudio.ai</a>. This Zenodo mirror has all audio task but only at 48000Hz sampling rate. For other sampling rates (16000, 22050, 32000, 44100), please download files (requester pays) from Google Storage gs://hear2021-archive/tasks/ or AWS s3://hear2021-archive/tasks/</p>
Dataset for the study "Changes in audio-spatial working memory abilities during childhood: The role of spatial and phonological development"
<p>Working memory is a cognitive system devoted to storage and retrieval processing of information.<br> Numerous studies on the development of working memory have investigated the<br> processing of visuo-spatial and verbal non-spatialized information; however, little is known<br> regarding the refinement of acoustic spatial and memory abilities across development.<br> Here, we hypothesize that audio-spatial memory skills improve over development, due to<br> strengthening spatial and cognitive skills such as semantic elaboration. We asked children<br> aged 6 to 11 years old (n = 55) to pair spatialized animal calls with the corresponding animal<br> spoken name. Spatialized sounds were emitted from an audio-haptic device, haptically<br> explored by children with the dominant hand’s index finger. Children younger than 8<br> anchored their exploration strategy on previously discovered sounds instead of holding this<br> information in working memory and performed worse than older peers when asked to pair<br> the spoken word with the corresponding animal call. In line with our hypothesis, these findings<br> demonstrate that age-related improvements in spatial exploration and verbal coding<br> memorization strategies affect how children learn and memorize items belonging to a complex<br> acoustic spatial layout. Similar to vision, audio-spatial memory abilities strongly depend<br> on cognitive development in early years of life.</p> <p>Data in the file are divided into six sheets based on the age of the participants and the experimental condition, either call-call or call-name. Each sheet contains six columns: Participant ID, age and gender are the first three. The last three columns instead refer to the test parameters: the number of attempts, the audio-anchor and the score. In details, the number of attempts indicates the number of trials needed to pair the sounds. The audio-anchor provides a measurement of the exploration strategy. It accounts for how many consecutive attempts the child begins by touching the same speaker while the score takes into account the frequency of touches on the same speakers: the more the participant returns on the same stimulus location, the lower the score.</p>
Tiny AAM: Sample from the Artificial Audio Multitracks Dataset
<p>This is the tiny demo version with 20 samples from the AAM dataset, which originally contains 3,000 artificial music audio tracks. In contrast to the full version, which has lossless audio, we only provide mp3 tracks for demonstration and testing purposes.</p> <p>For the <strong>full version</strong> and <strong>more information</strong> see: <a href="https://doi.org/10.5281/zenodo.5794629">zenodo.5794629</a></p>
AcousticsAustraliaArticle_dataset: audio samples _ vowel phonation
<p><strong>PaperAcousticsAustralia - data</strong></p> <p> </p> <p>Audio samples were collected from 31 participants, extracted from their recorded reading (31 participants), recorded interviews (28 participants) and vowel phonations (29 participants).</p> <p><strong>ParticipantProfiles spreadsheet</strong> details the age, gender, years smoking and ethnicity per participant.</p> <p> <strong>ReadingsAudioSamples : </strong>audio samples extracted from recorded readings. There are generally two reading audio samples per participant. The measures reported per participant are the mean values read from PRAAT for the two samples extracted for each participant.</p> <p> <strong>InterviewsAudioSamples :</strong> audio samples extracted from recorded interviews. This concerns 28 participants: there is no audio samples available from the interviews of participants P20, P33 and P42. There are generally at least two audio samples per participant. The measures reported per participant are the mean values read from PRAAT for the participant’s audio samples.</p> <p> <strong>VowelPhonationAudioSamples :</strong> audio samples of vowel /a/ sustained phonation for 29 participants. Most samples were recorded by the participants, and eight of them were recorded by phone (look for ‘phone’ in the audio sample name). There is only one vowel phonation sample per participant and no vowel phonation sample was provided for participants P16 and P25. </p>
AcousticsAustraliaArticle_dataset: audio samples _ readings
<p><strong>PaperAcousticsAustralia - data</strong></p> <p> </p> <p>Audio samples were collected from 31 participants, extracted from their recorded reading (31 participants), recorded interviews (28 participants) and vowel phonations (29 participants).</p> <p><strong>ParticipantProfiles spreadsheet</strong> details the age, gender, years smoking and ethnicity per participant.</p> <p> <strong>ReadingsAudioSamples : </strong>audio samples extracted from recorded readings. There are generally two reading audio samples per participant. The measures reported per participant are the mean values read from PRAAT for the two samples extracted for each participant.</p> <p> <strong>InterviewsAudioSamples :</strong> audio samples extracted from recorded interviews. This concerns 28 participants: there is no audio samples available from the interviews of participants P20, P33 and P42. There are generally at least two audio samples per participant. The measures reported per participant are the mean values read from PRAAT for the participant’s audio samples.</p> <p> <strong>VowelPhonationAudioSamples :</strong> audio samples of vowel /a/ sustained phonation for 29 participants. Most samples were recorded by the participants, and eight of them were recorded by phone (look for ‘phone’ in the audio sample name). There is only one vowel phonation sample per participant and no vowel phonation sample was provided for participants P16 and P25. </p>
AcousticsAustraliaArticle_dataset: audio samples _ Interviews
<p><strong>PaperAcousticsAustralia - data description</strong></p> <p> </p> <p>Audio samples were collected from 31 participants, extracted from their recorded reading (31 participants), recorded interviews (28 participants) and vowel phonations (29 participants).</p> <p><strong>ParticipantProfiles spreadsheet</strong> details the age, gender, years smoking and ethnicity per participant.</p> <p> <strong>ReadingsAudioSamples: </strong>audio samples extracted from recorded readings. There are generally two reading audio samples per participant. The measures reported per participant are the mean values read from PRAAT for the two samples extracted for each participant.</p> <p> <strong>InterviewsAudioSamples:</strong> audio samples extracted from recorded interviews. This concerns 28 participants: there is no audio samples available from the interviews of participants P20, P33 and P42. There are generally at least two audio samples per participant. The measures reported per participant are the mean values read from PRAAT for the participant’s audio samples.</p> <p> <strong>VowelPhonationAudioSamples: </strong>audio samples of vowel /a/ sustained phonation for 29 participants. Most samples were recorded by the participants, and eight of them were recorded by phone (look for ‘phone’ in the audio sample name). There is only one vowel phonation sample per participant and no vowel phonation sample was provided for participants P16 and P25. </p> <p> </p> <p> </p>
JalCommunDisorderAssistTechnol_dataset: audio samples_interviews
<p><strong>PaperJalCommunDisorderAssistTechnol - data</strong></p> <p> </p> <p>Audio samples were collected from 59 participants, either extracted from recorded reading (34 participants) or from a recorded interview (the remaining 25 participants).</p> <p><strong>ParticipantProfiles spreadsheet </strong>details the age, gender, and years smoking per participant.</p> <p> <strong>ReadingsAudioSamples :</strong> audio samples extracted from recorded readings. This concerns 34 participants. There are generally two reading audio samples per participant. The measures reported per participant are the mean values read from PRAAT for the two samples per participant.</p> <p> <strong>InterviewsAudioSamples :</strong> audio samples extracted from recorded interviews. This concerns 25 participants. There is generally only one audio sample per participant. When there is more than one audio sample for a same participant, the measures reported for the participant are the mean values read from PRAAT for the participant’s audio samples.</p>
JalCommunDisorderAssistTechnol_dataset: audio samples_readings
<p><strong>PaperJalCommunDisorderAssistTechnol - data</strong></p> <p> </p> <p>Audio samples were collected from 59 participants, either extracted from recorded reading (34 participants) or from a recorded interview (the remaining 25 participants).</p> <p><strong>ParticipantProfiles spreadsheet </strong>details the age, gender, and years smoking per participant.</p> <p> <strong>ReadingsAudioSamples :</strong> audio samples extracted from recorded readings. This concerns 34 participants. There are generally two reading audio samples per participant. The measures reported per participant are the mean values read from PRAAT for the two samples per participant.</p> <p> <strong>InterviewsAudioSamples :</strong> audio samples extracted from recorded interviews. This concerns 25 participants. There is generally only one audio sample per participant. When there is more than one audio sample for a same participant, the measures reported for the participant are the mean values read from PRAAT for the participant’s audio samples.</p>
Head-motion and eye-gaze behavior reveal audio-visual target search strategies - dataset
<p>Participants were tasked with finding a target stimulus in the presence of a number of auditory distractors. </p> <p>The target stimulus (audio-only, audio-visual or visual-only) was presented at the central position for 3 seconds, after which the stimulus was moved to one of 24 positions around the participant. The other 23 positions all contained a visual distractor and 0, 1, 2, 3, 5, 7 or 11 auditory distractors (evenly spaced). </p> <p>Based on the eye and headtracking data, we calculated the FOV and target localization time and the maximum headrotation into the wrong direction. </p> <p>FOV localization time: time it took to bring target within FOV. <br>Target localization time: From FOV to response.</p> <p>These datasets contain both the tracking data and the summarized data. </p> <p>Columns "av_stim_x_y" list for each of the 24 AV stimuli (which both served as the targets and distractors), the id, the angle at which it was present during the trial and the visual and audio status.</p> <p>FUNDING: </p> <p>The research was supported by the Centre for Applied Hearing<br>research (CAHR) through a research consortium agreement with<br>GN Resound, Oticon, and Widex. The funders had no role in<br>study design, data collection and analysis, decision to publish, or<br>preparation of the article.</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.