Subject: Re: 60 Hz hum causing problems with speech recognition From: Tarun Pruthi <t.pruthi@xxxxxxxx> Date: Tue, 16 Jun 2009 13:01:32 -0400 List-Archive:<http://lists.mcgill.ca/scripts/wa.exe?LIST=AUDITORY>--001636c5ac7a66fdc1046c7a20f4 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Hi Linda: I am trying to make a speech command recognizer using a standard open-source speech recognizer on my laptop. Most of the work I do, is either on my office desk or home. I guess the sources of noise would be the usual things: lights, unclean power, wireless signals etc. Tarun On Tue, Jun 16, 2009 at 11:57 AM, Linda Seltzer <lseltzer@xxxxxxxx > wrote: > What exactly is the experiment you are trying to do or the test that you > are conducting and where is the recording being done? > > Hi all: > > I have been having a lot of problems lately with 60 Hz humming noise in > my > > speech recordings (I will confess that most of my tests have been on > > windows/linux laptops - and my guess is that they probably have much > worse > > recording quality than desktops on average - but we live in a mobile > world > > now and I have to make it work on laptops). Whenever it is there, it > > causes > > a serious drop in speech recognition accuracy, especially since it comes > > in > > very randomly at times giving for example clean recordings during speaker > > enrollment and noisy recordings in live testing. The problem is even > worse > > on linux which has given me all kinds of artifacts in the recordings > > besides > > the 60 Hz hum - my guess is that this is most likely because of driver > > problems. > > > > So, I have 3 questions: > > 1. What is the source of this 60 Hz noise? Powerline? RF interference? I > > have tried disconnecting the power cord - it works sometimes, but not > > always. > > 2. Is there a standard way to get rid of the 60 Hz hum and ensure clean > > recordings for the purposes of speech recognition? The only way I can > > think > > of is to use notch filters at 60 Hz, and some of the harmonics, say 120 > Hz > > and 180 Hz, since most of the times it doesn't manifest as a single tone > > at > > 60 Hz, but has harmonics all over the frequency range. > > 3. If this is a well-known problem with sound cards, then why aren't we, > > speech recognition community, trying to come up with a mandatory > > compliance > > standard for sound card manufacterers which ensures that the microphone > > input is not corrupted by 60/50 Hz noise? That should provide a big boost > > to > > speech recognition accuracies, and a wider acceptance of speech > > recognition. > > > > Tarun > > Senior Research Engineer > > Think A Move, Ltd > > > > --001636c5ac7a66fdc1046c7a20f4 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Hi Linda:<div><br></div><div>I am trying to make a speech command recognize= r using a standard open-source speech recognizer on my laptop. Most of the = work I do, is either on my office desk or home. I guess the sources of nois= e would be the usual things: lights, unclean power, wireless signals etc.</= div> <div><br></div><div>Tarun<br><br><div class=3D"gmail_quote">On Tue, Jun 16,= 2009 at 11:57 AM, Linda Seltzer <span dir=3D"ltr"><<a href=3D"mailto:ls= eltzer@xxxxxxxx">lseltzer@xxxxxxxx</a>></span> wrote= :<br> <blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p= x #ccc solid;padding-left:1ex;">What exactly is the experiment you are tryi= ng to do or the test that you<br> are conducting and where is the recording being done?<br> <div><div></div><div class=3D"h5">> Hi all:<br> > I have been having a lot of problems lately with 60 Hz humming noise i= n my<br> > speech recordings (I will confess that most of my tests have been on<b= r> > windows/linux laptops - and my guess is that they probably have much w= orse<br> > recording quality than desktops on average - but we live in a mobile w= orld<br> > now and I have to make it work on laptops). Whenever it is there, it<b= r> > causes<br> > a serious drop in speech recognition accuracy, especially since it com= es<br> > in<br> > very randomly at times giving for example clean recordings during spea= ker<br> > enrollment and noisy recordings in live testing. The problem is even w= orse<br> > on linux which has given me all kinds of artifacts in the recordings<b= r> > besides<br> > the 60 Hz hum - my guess is that this is most likely because of driver= <br> > problems.<br> ><br> > So, I have 3 questions:<br> > 1. What is the source of this 60 Hz noise? Powerline? RF interference?= I<br> > have tried disconnecting the power cord - it works sometimes, but not<= br> > always.<br> > 2. Is there a standard way to get rid of the 60 Hz hum and ensure clea= n<br> > recordings for the purposes of speech recognition? The only way I can<= br> > think<br> > of is to use notch filters at 60 Hz, and some of the harmonics, say 12= 0 Hz<br> > and 180 Hz, since most of the times it doesn't manifest as a singl= e tone<br> > at<br> > 60 Hz, but has harmonics all over the frequency range.<br> > 3. If this is a well-known problem with sound cards, then why aren'= ;t we,<br> > speech recognition community, trying to come up with a mandatory<br> > compliance<br> > standard for sound card manufacterers which ensures that the microphon= e<br> > input is not corrupted by 60/50 Hz noise? That should provide a big bo= ost<br> > to<br> > speech recognition accuracies, and a wider acceptance of speech<br> > recognition.<br> ><br> > Tarun<br> > Senior Research Engineer<br> > Think A Move, Ltd<br> ><br> <br> </div></div></blockquote></div><br></div> --001636c5ac7a66fdc1046c7a20f4--