Subject: [AUDITORY] Track separation algorithm From: Sharath Chandra Ram <looplogic@xxxxxxxx> Date: Sat, 29 Jun 2024 10:59:09 -0400--0000000000003b0dc2061c089b21 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable You might want to consider Spleeter (uses pre-trained tensor flow models , and you can train your own if you have data) https://github.com/deezer/spleeter You can =E2=80=98pip install spleeter=E2=80=99 https://pypi.org/project/spl= eeter/ Many commercial solutions including one for Max/Ableton have been forked off Spleeter. On Friday, June 28, 2024, Jeremie Ginzburg <jeremie.ginzburg@xxxxxxxx> wrote: > Dear Auditory-listers, > > We are trying to find a way to separate speech and music (and other track= s > if possible) from a movie audio track. > > We are not aware of the latest advances in that specific area. Is anyone > aware of (maybe AI-based) tools to do that? And if so, what is your opini= on > on its reliability? > > Thank you and have a nice weekend :) > > -- > J=C3=A9r=C3=A9mie Ginzburg > Post-doctoral Researcher - McGill University > --=20 Sharath Chandra Ramakrishnan (M.S., Ph.D.) Assistant Professor University of Illinois Urbana-Champaign --0000000000003b0dc2061c089b21 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable You might want to consider Spleeter (uses pre-trained tensor flow models , = and you can train your own if you have data)<div><a href=3D"https://github.= com/deezer/spleeter" target=3D"_blank">https://github.com/deezer/sple<wbr>e= ter</a></div><div><br></div><div>You can =E2=80=98pip install spleeter=E2= =80=99=C2=A0<a href=3D"https://pypi.org/project/spleeter/" target=3D"_blank= ">https://pypi.org/<wbr>project/spleeter/</a></div><div><br></div><div>Many= commercial solutions including one for Max/Ableton have been forked off Sp= leeter.</div><div><br></div><div><br>On Friday, June 28, 2024, Jeremie Ginz= burg <<a href=3D"mailto:jeremie.ginzburg@xxxxxxxx" target=3D"_blank">je= remie.ginzburg@xxxxxxxx</a>> wrote:<br><blockquote class=3D"gmail_quote= " style=3D"margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">D= ear Auditory-listers,<br> <br> We are trying to find a way to separate speech and music (and other tracks = if possible) from a movie audio track.<br> <br> =C2=A0We are not aware of the latest advances in that specific area. Is any= one aware of (maybe AI-based) tools to do that? And if so, what is your opi= nion on its reliability?<br> <br> Thank you and have a nice weekend :)<br> <br> -- <br> J=C3=A9r=C3=A9mie Ginzburg<br> Post-doctoral Researcher - McGill University<br> </blockquote></div> <br><br>-- <br>Sharath Chandra Ramakrishnan <br>(M.S., Ph.D.)<br>Assistant = Professor<br>University of Illinois Urbana-Champaign<p><p><p><br><br></p></= p></p> --0000000000003b0dc2061c089b21--