PNCC.INTERSPEECH2009
所属分类:matlab编程
开发工具:matlab
文件大小:520KB
下载次数:31
上传日期:2012-03-15 17:49:48
上 传 者:
am3roch3
说明: PNCCs by C.Kim code sample
文件列表:
NormalizeFilterGain.m (363, 2009-04-12)
PNCC.m (9693, 2009-04-19)
ComputeFilterResponse.m (7712, 2009-04-12)
SpectrogramDemo\ComputeFilterResponse.m (7712, 2009-04-12)
SpectrogramDemo\DemoBatch.asv (3014, 2009-04-19)
SpectrogramDemo\DemoBatch.m (3105, 2009-04-19)
SpectrogramDemo\NormalizeFilterGain.m (363, 2009-04-12)
SpectrogramDemo\PNCCForDemo.asv (8955, 2009-04-19)
SpectrogramDemo\PNCCForDemo.m (9699, 2009-04-19)
SpectrogramDemo\sb01_Clean.sph (165888, 2007-08-30)
SpectrogramDemo\sb01_Music_05dB.sph (165888, 2007-06-28)
SpectrogramDemo\sb01_Street_05dB.sph (165888, 2008-09-08)
SpectrogramDemo\sb01_White_05dB.sph (165888, 2007-06-28)
Programmed by Chanwoo Kim
for Interspeech 2009
Apr 10, 2009
1) Just Run PNCC('outFile', 'inPutFile')
IMPORTANT : The input is assumed in the single-Channel sphere NIST format.
The sampling rate
should be "16 kHz".
It does not check the header, but just skips the header.
We used this program in getting result in INTERSPEECH 2009 paper on the RM1 database
2) Output is a feature in Sphinx format
3) To see the spectrogram demo
Go to SpectrogramDemo directory
Type DemoBatch.m in Matlab
It will launch three figures:
a) spectrogram,
b) spectrogram using log nonliearity (This one is still better than MFCC in our recognition experiment)
c) gammatone weighting, PNCC (This one is much better than b))
In each figure you will sub-figures obtained from different noisy conditions. If the spectrogram from the noisy
condition is closer to the top subfigure in each figure, then we can think that this feature is more robust.
近期下载者:
相关文件:
收藏者: