iiscleap / coswara-data Goto Github PK

View Code? Open in Web Editor NEW

175.0 21.0 147.0 15.98 GB

Data repository of Project Coswara

Home Page: https://coswara.iisc.ac.in

License: Other

Python 15.79% Shell 6.53% Jupyter Notebook 77.68%

cough covid-19 coronavirus machine-learning

coswara-data's Introduction

Coswara-Data

This work is licensed under a Creative Commons Attribution 4.0 International (CC BY 4.0) License.

UPDATE: The current full version of Coswara data is now published with open access in Nature Scientific Data, 2023. read

Project Coswara by Indian Institute of Science (IISc) Bangalore is an attempt to build a diagnostic tool for COVID-19 detection using the audio recordings such as breathing, cough and speech sounds of an individual. Currently, the project is in the data collection stage through crowdsourcing. To contribute your audio samples, please go to Project Coswara(https://coswara.iisc.ac.in/). The exercise takes 5-7 minutes.

What am I looking at? This github repository contains the raw audio data collected through https://coswara.iisc.ac.in/ . Every participant contributes nine sound samples. You can read the paper: Coswara - A Database of Breathing, Cough, and Voice Sounds for COVID-19 Diagnosis to know more about the dataset. Note that the dataset size has increased since this paper came out. We also maintain a (less frequently updated) blog here.

What is the structure of the repository? Each folder contains metadata and audio recordings corresponding to contributors. The folder is compressed. To download and extract the data, you can run the script extract_data.py

What are the different sound samples? Sound samples collected include breathing sounds (fast and slow), cough sounds (deep and shallow), phonation of sustained vowels (/a/ as in made, /i/,/o/), and counting numbers at slow and fast pace. Metadata information collected includes the participant's age, gender, location (country, state/ province), current health status (healthy/ exposed/ positive/recovered) and the presence of comorbidities (pre-existing medical conditions).

Can I see the metadata before downloading whole repository? Yes. The file combined_data.csv contains a summary of metadata. The file csv_labels_legend.json contains information about the columns present in combined_data.csv.

Is there any audio quality check? Yes. The audio files are manually listened and labeled as one of the three categories: 2(excellent), 1(good), 0(bad). The labels are present in the annotations folder.

How to cite this dataset in your work? Great to know you found it useful. You can cite the paper: Coswara - A Database of Breathing, Cough, and Voice Sounds for COVID-19 Diagnosis (https://arxiv.org/abs/2005.10548)

Is there any web application for COVID-19 screening based on respiratory acoustics? Yes! One can record his/her respiratory sounds at Coswara web application and obtain a COVID-19 probability score in few seconds. Demo: here, paper: here

What is the count of participants in each folder?

2020-04-13 contains 76 samples.
2020-04-15 contains 161 samples.
2020-04-16 contains 197 samples.
2020-04-17 contains 168 samples.
2020-04-18 contains 46 samples.
2020-04-19 contains 32 samples.
2020-04-24 contains 28 samples.
2020-04-30 contains 23 samples.
2020-05-02 contains 155 samples.
2020-05-04 contains 81 samples.
2020-05-05 contains 14 samples.
2020-05-25 contains 54 samples.
2020-06-04 contains 20 samples.
2020-07-07 contains 42 samples.
2020-07-20 contains 21 samples.
2020-08-03 contains 29 samples.
2020-08-14 contains 83 samples.
2020-08-20 contains 48 samples.
2020-08-24 contains 19 samples.
2020-09-01 contains 24 samples.
2020-09-11 contains 16 samples.
2020-09-19 contains 32 samples.
2020-09-30 contains 26 samples.
2020-10-12 contains 18 samples.
2020-10-31 contains 29 samples.
2020-11-30 contains 17 samples.
2020-12-21 contains 27 samples.
2021-02-06 contains 18 samples.
2021-04-06 contains 66 samples.
2021-04-19 contains 35 samples.
2021-04-26 contains 41 samples.
2021-05-07 contains 54 samples.
2021-05-23 contains 31 samples.
2021-06-03 contains 42 samples.
2021-06-18 contains 56 samples.
2021-06-30 contains 67 samples.
2021-07-14 contains 52 samples.
2021-08-16 contains 82 samples.
2021-08-30 contains 64 samples.
2021-09-14 contains 37 samples.
2021-09-30 contains 103 samples.
2022-01-16 contains 141 samples.
2022-02-24 contains 372 samples.

Each folder also has a CSV file which contains metadata of each sample (that is, participant).

Can I know the individuals maintaining this project? Yes, we are a team of Professors, PostDocs, Engineers, and Research Scholars affiliated with the Indian Institute of Science, Bangalore (India). Sriram Ganapathy, Assistant Professor, Dept. Electrical Engineering, IISc is the Principal Investigator of this project.

Current Members: Debarpan Bhattacharya, Neeraj Kumar Sharma, Prasanta Kumar Ghosh, Srikanth Raj Chetupalli, Sriram Ganapathy

Past Members: Anand Mohan, Ananya Muguli, Debottam Dutta, Prashant Krishnan, Pravin Mote, Rohit Kumar, Shreyas Ramoji

(arranged in alphabetical order)

coswara-data's People

Contributors

Stargazers

Watchers

Forkers

karanalytics abhinav43 drewgrant aascode hkp2k19 bloodcarter leosaiyanx janashreea ankitshah009 aniashok bhanudaybirla ananyamuguli sanket7738 kedarzope himanshumangal09 sourabh-714 itsdandelia saisriram1 sanika-thakur itzzdeep shreeshan betty5257 raaghavrm rashy98 koopa01 avakshaya1996 xiaocuilin rodcanada luissalvadorheysen chandrahasahu chang111 shivangtiwari tfg-anonimo qshobak arup94124 mstfakdgn zenodia gvaladao muyun-li xonlin vishwas-yogi shreyasdesikan zzhao38 rajeshsrivathsa dgsivan beyzant blesslord theghostsenpai himanshukgp007 dkinopoulos azam64064 aissa94hadji thuy4tbn99 pranaynandan berong91 nt-thuhang jeetyadav sganguly1999 iam-pattan jiwang176 mostafa54225 arulananthamanujan nayoumi akangkshyap anushkadwivedi17 bamblebam prateek10599 madhu-baala jokerinfini hemendrasharma04 srikeshram oyeeabhishek iot08 mbencherif loganliu66 aaaeeee shen-jk porjaiactive pramestia ameyasm1154 binhfdv kraketito hado2020 tavneetgill toducthanh isfandyar haiyinxuan cyberflamego ezraorich yasserelmenshawy kalpana200 ranjana-kambhammettu janithdesilva xiaoyeasz yaswanthi01 vsdcjs dvoynikova sbrichardson elizabethcherian lavanya-cs-ui

coswara-data's Issues

Problems with covid_status column

I'm having some problems to understand the meaning of the labels in covid_satus column.

What does mean "resp_illness_not_identified" and "no_resp_illness_exposed"?

"resp_illness_not_identified" = the person does not have respiratory problems? or the person has respiratory problems but dont know what illness he has?
"no_resp_illness_exposed" = the person does not have respiratory problems but he was exposed to corona virus?

Data is corrupted, Not able to open it.

I tried to open the tar file using this command

tar -xvzf 20200415.tar.gz.aa

I am getting this error :

x 20200415/rSDhZrvtPFhvXnAPUJnYoPBN0mD3/counting-normal.wav: (Empty error message)
tar: Error exit delayed from previous errors.

Paper mentioned there are 3470 clean audios, but the sample count from README add up to less than 1500 samples

Same as title. Any clarification ?

Paper link: https://arxiv.org/abs/2005.10548

I do not open files named tar.gz.ab-ac-ad..aj, but the file named tar.gz.aa can be opened. There is no problem with the files named aa. How can I solve this problem?

The terminal error is as follows;
tar: Error opening archive: Unrecognized archive format
When I investigated the terminal error, I could not reach the results to solve the error. I am a macOS user. How can I solve this problem?
I would be very grateful if you could help.

data extract error

I run extract_data.py . it create a folder Extracted_data . but is empty! . program run by environment win10, pycharm

Clone the repo

It seems that the provided link for cloning the repo is not a valid one. my terminal says that it is not a repo. (.git) is missing, isn't it?

Unable to Open

I am not able to access the tar files, as they are in patterns of aa, ab and so on

Missing id numbers in the outcome variable file

Thank you for this great dataset. Will the ids for the most recent set of participants (pushed 07/07) be paired with the provided recordings? The ids are currently missing from the .csv file. Thank you!

unable to extract in windows

The archive is corrupt

Issue regarding extracting the dataset

I have run the extract_data.py file but it could download the wav file , it just create a empty folder called extracted data , kindly help me out thanks

Extracting data windows

When I try to run the extract data script on windows I get a message saying <'cat' is not recognized as an internal or external command> and no data is extracted. Do you have a workaround?

Unrecognized format error message when extracting ".ab", ".ac", ".ad", etc.

Hello! Thanks for the dataset! I have been having quite a bit of trouble extracting the tar files unless they have the ".aa" extension. The other tar files (e.g. ".ab", ".ac", ".ad", etc.) give my tar extractors error messages suggesting the tar files are not archives and are in an unrecognized format. Could someone please help me figure out how to resolve this issue?

gzip: stdin: unexpected end of file tar: Child returned status 1 tar: Error is not recoverable: exiting now The syntax of the command is incorrect.

I am getting this error when I am running the extract_data.py script.

How to extract .aa files & How to split the file for each patient

Hi Team,

I am facing issue in using the .aa files in each date dataset , Need your assistance in extract the data

Question 1 :- How we can extract the .aa files , Is there any python Code snippet or library exist

Question 2 :-
I have came across "from pydub import AudioSegment" library , What is the params for each patient
frame_rate
channels
sample_width

Each .aa files definitely contains samples of multiple patient / day

Thanks

The full name of label in CSV file

Actually, I cannot find the full name of labels in CSV file, such as ep, rU, ht, um, ihd and etc, though some labels are obviously.
Could you please upload a file like a map, a -- age, g--gender. Thank you.