[AUDITORY] Announcing SOUNDATA: A Python library for reproducible use of audio datasets (Justin Salamon )

Subject: [AUDITORY] Announcing SOUNDATA: A Python library for reproducible use of audio datasets From: Justin Salamon <000000b4a42fd03d-dmarc-request@xxxxxxxx> Date: Wed, 3 Nov 2021 00:26:38 +0000 --_000_BYAPR02MB53333ED80753C5198BD5FEC5AB8C9BYAPR02MB5333namp_ Content-Type: text/plain; charset="Windows-1252" Content-Transfer-Encoding: quoted-printable *** apologies for any cross-postings *** Dear colleagues, We=92re excited to announce the release of soundata, a python library for r= eproducible use of audio datasets. Soundata can be installed via: pip install soundata The source code lives here: https://github.com/soundata/soundata We=92re launching with 14 popular environmental sound datasets<https://soun= data.readthedocs.io/en/latest/source/quick_reference.html>, with plans to c= ontinue expanding with additional datasets spanning a range of audio domain= s including speech and bioacoustics. For music datasets see mirdata<https:/= /github.com/mir-dataset-loaders/mirdata>, which was the inspiration for sou= ndata. Soundata makes it easy to: * Download datasets to a common location and format * Validate that a downloaded dataset is complete and perfectly matches = a canonical version * Load audio and annotation files into a common format * Parse clip-level metadata for detailed evaluations We hope soundata will help the community to: * Ensure results are reproducible by working against exactly the same d= ata * Save time by avoiding manual downloads and having to write custom dat= aset parsers * Automate large-scale download, training, and evaluation pipelines * Increase the visibility of new datasets by adding them to soundata Soundata is a cross-organizational collaboration spanning researchers from = MARL@xxxxxxxx<https://steinhardt.nyu.edu/marl>, Adobe Research<https://research.= adobe.com/research/audio/>, MTG@xxxxxxxx<https://www.upf.edu/web/mtg>, and GPA@xxxxxxxx= delaR<https://iie.fing.edu.uy/investigacion/grupos/gpa/en/audio-processing-= group/>. You can learn more about the library on our docs page: https://soundata.rea= dthedocs.io/ A bit more about the motivation for soundata can be found in our (work in p= rogress) paper: "Soundata: A Python library for reproducible use of audio datasets" Magdalena Fuentes, Justin Salamon, Pablo Zinemanas, Mart=EDn Rocamora, Gen= =EDs Plaja, Ir=E1n R. Rom=E1n, Marius Miron, Xavier Serra, Juan Pablo Bello [arXiv<https://arxiv.org/abs/2109.12690>] We *welcome and encourage* contributions from the community, especially dat= a loaders for datasets not included yet in soundata. Cheers, Justin & Magdalena on behalf of the soundata team -- Justin Salamon | Adobe Research | www.justinsalamon.com --_000_BYAPR02MB53333ED80753C5198BD5FEC5AB8C9BYAPR02MB5333namp_ Content-Type: text/html; charset="Windows-1252" Content-Transfer-Encoding: quoted-printable <html> <head> <meta http-equiv=3D"Content-Type" content=3D"text/html; charset=3DWindows-1= 252"> <style type=3D"text/css" style=3D"display:none;"> P {margin-top:0;margin-bo= ttom:0;} </style> </head> <body dir=3D"ltr"> <div style=3D"font-family: Arial, Helvetica, sans-serif; font-size: 10pt; c= olor: rgb(0, 0, 0);"> *** apologies for any cross-postings *** Dear colleagues, We=92re excited to announce the release of&nbsp= ;soundata, a python library for reproducible use of audio datasets. Soundata can be installed via: </sp= an>pip install soundata The source code lives here: = <a href=3D"https://github.com/soundata/soundata" style=3D"margin:0px"><span= style=3D"margin:0px;font-weight:400;font-size:11pt;font-family:Arial;color= :rgb(17, 85, 204);text-decoration:underline;text-decoration-skip-ink:none">= https://github.com/soundata/soundata</a> We=92re launching with <a hr= ef=3D"https://soundata.readthedocs.io/en/latest/source/quick_reference.html= " style=3D"margin:0px">14 popular environmental sound datasets</a>, with plans to continue e= xpanding with additional datasets spanning a range of audio domains includi= ng speech and bioacoustics. For music datasets see <a href=3D"http= s://github.com/mir-dataset-loaders/mirdata" style=3D"margin:0px">mirda= ta</a>, which was the inspiration for soundata. Soundata makes it easy to: <ul style=3D"background-color:rgb(255, 255, 255);font-weight:bold;margin-to= p:0px;margin-bottom:0px;padding-inline-start:48px"> <li dir=3D"ltr" style=3D"font-size:11pt;font-family:Arial;font-weight:400"> = Download datasets to a common location and forma= t </li><li dir=3D"ltr" style=3D"font-size:11pt;font-family:Arial;font-weight:= 400"> = Validate that a downloaded dataset is complete a= nd perfectly matches a canonical version </li><li dir=3D"ltr" style=3D"font-size:11pt;font-family:Arial;font-weight:= 400"> = Load audio and annotation files into a common fo= rmat </li><li dir=3D"ltr" style=3D"font-size:11pt;font-family:Arial;font-weight:= 400"> = Parse clip-level metadata for detailed evaluatio= ns </li></ul> We hope soundata will help the community to: <ul style=3D"background-color:rgb(255, 255, 255);font-weight:bold;margin-to= p:0px;margin-bottom:0px;padding-inline-start:48px"> <li dir=3D"ltr" style=3D"font-size:11pt;font-family:Arial;font-weight:400"> = Ensure results are reproducible by working again= st exactly the same data </li><li dir=3D"ltr" style=3D"font-size:11pt;font-family:Arial;font-weight:= 400"> = Save time by avoiding manual downloads and havin= g to write custom dataset parsers </li><li dir=3D"ltr" style=3D"font-size:11pt;font-family:Arial;font-weight:= 400"> = Automate large-scale download, training, and eva= luation pipelines </li><li dir=3D"ltr" style=3D"font-size:11pt;font-family:Arial;font-weight:= 400"> = Increase the visibility of new datasets by addin= g them to soundata </li></ul> Soundata is a cross-organizational collaboration spanning researchers from=  <a href=3D"https://steinhard= t.nyu.edu/marl" style=3D"margin:0px">MARL@xxxxxxxx</a>, <a href=3D"https://research.adobe.com/r= esearch/audio/" style=3D"margin:0px">Adobe Research</a>, <a hre= f=3D"https://www.upf.edu/web/mtg" style=3D"margin:0px">MTG@xxxxxxxx<= /a>, and <a href=3D"https://iie.f= ing.edu.uy/investigacion/grupos/gpa/en/audio-processing-group/" style=3D"ma= rgin:0px">GPA@xxxxxxxx</a>.  You can learn more about the library on our docs page: <a href=3D"https://soundata.readthedocs.io/" sty= le=3D"margin:0px">https://soundata.readthedocs.io/</a> A bit more about the motivation for soundata can be found in our (work in = progress) paper: "Soundata:= A Python library for reproducible use of audio datasets" Magdalena Fuentes, Justin Salamon, Pablo Zinemanas, Mart=EDn Rocamora, Gen= =EDs Plaja, Ir=E1n R. Rom=E1n, Marius Miron, Xavier Serra, Juan Pablo Bello= [<a href=3D"https://arxiv.org/abs/2109.12690" style=3D"margin:0px">= arXiv</a>] We *welcome and encourage* contributions from the community, especially da= ta loaders for datasets not included yet in soundata. Cheers, Justin & Magdalena on behalf of the soundata team </div> <div> <div style=3D"font-family: Arial, Helvetica, sans-serif; font-size: 10pt; c= olor: rgb(0, 0, 0);"> </div> <div id=3D"Signature"> <div> <div style=3D"font-family:Arial,Helvetica,sans-serif; font-size:10pt; color= :rgb(0,0,0)"> --</div> <div style=3D"font-family:Arial,Helvetica,sans-serif; font-size:10pt; color= :rgb(0,0,0)"> Justin Salamon | Adobe Research | www.justinsalamon.com</div> </div> </div> </div> </body> </html> --_000_BYAPR02MB53333ED80753C5198BD5FEC5AB8C9BYAPR02MB5333namp_--

This message came from the mail archive
src/postings/2021/
maintained by:

DAn Ellis <dpwe@ee.columbia.edu>
Electrical Engineering Dept., Columbia University