help in speaker segmentation
in [Matlab]
|
From: lama on 16 Jun 2010 10:28 hi, I am new in matlab and I have the time boundaries for each speaker in the audio file. How I will treat this as segments to do the clustering procedures? Thanks, Lama
From: us on 16 Jun 2010 10:43 "lama " <radafn(a)gmail.com> wrote in message <hvan1n$868$1(a)fred.mathworks.com>... > hi, > I am new in matlab and I have the time boundaries for each speaker in the audio file. How I will treat this as segments to do the clustering procedures? > > Thanks, > Lama a hint: - split the input array into smaller segments (vecs) according to the time boundaries, which you seem to know... us
From: lama on 16 Jun 2010 10:50 "us " <us(a)neurol.unizh.ch> wrote in message <hvanto$56g$1(a)fred.mathworks.com>... > "lama " <radafn(a)gmail.com> wrote in message <hvan1n$868$1(a)fred.mathworks.com>... > > hi, > > I am new in matlab and I have the time boundaries for each speaker in the audio file. How I will treat this as segments to do the clustering procedures? > > > > Thanks, > > Lama > > a hint: > - split the input array into smaller segments (vecs) according to the time > boundaries, which you seem to know... > > us Thanks for ur respond. So do u mean I have to split the audio according the time boundaries manually
From: us on 16 Jun 2010 11:02 "lama " <radafn(a)gmail.com> wrote in message <hvaobg$2vc$1(a)fred.mathworks.com>... > "us " <us(a)neurol.unizh.ch> wrote in message <hvanto$56g$1(a)fred.mathworks.com>... > > "lama " <radafn(a)gmail.com> wrote in message <hvan1n$868$1(a)fred.mathworks.com>... > > > hi, > > > I am new in matlab and I have the time boundaries for each speaker in the audio file. How I will treat this as segments to do the clustering procedures? > > > > > > Thanks, > > > Lama > > > > a hint: > > - split the input array into smaller segments (vecs) according to the time > > boundaries, which you seem to know... > > > > us > > Thanks for ur respond. > > So do u mean I have to split the audio according the time boundaries manually well... you COULD(!) manually write a small code snippet (ML-function/-script), which cycles through your time-list, cuts the segments for you, and puts them into a CELL or a numerical array... us
From: ImageAnalyst on 16 Jun 2010 11:10
That is: audioForSpeaker1 = allSpeakersAudio(speaker1Start:speaker1Stop); audioForSpeaker2 = allSpeakersAudio(speaker2Start:speaker2Stop); audioForSpeaker3 = allSpeakersAudio(speaker3Start:speaker3Stop); etc. and so on for all the other speakers. You said "I have the time boundaries for each speaker " so that means you have speaker1Start, speaker1Stop, speaker2Start, speaker2Stop, etc. |