Sign in

username:

password:



Not a member?

Search compdsp



Search tips

comp.dsp by Keywords

Adaptive Filter | ADPCM | ADSP | ADSP-2181 | Aliasing | AMR | Anti-Aliasing | ARMA | Autocorrelation | AutoCovariance | Beamforming | Bessel | Blackfin | Butterworth | C6713 | CCS | Chebyshev | CIC Filter | Circular Convolution | Code Composer Studio | Comb Filter | Compression | Convolution | Cross Correlation | DCT | Decimation | Deconvolution | Demodulation | DM642 | DSP Boards | DSP/BIOS | DTMF | Echo Cancellation | Equalization | Equalizer | ETSI | EZLITE (Ez-kit Lite) | FFT | FFTW | FIR Filter | Fixed Point | FSK | G.711 | G.723 | G.729 | Gaussian Noise | Goertzel | GPIO | Hilbert Transform | IFFT | IIR Filter | Interpolation | Invariance | JTAG | Kalman | Laplace Transform | Levinson | LPC | McBSP | MIPS | Modulation | MPEG | Multirate | Notch Filter | Nyquist | OFDM | Oversampling | Pink Noise | Pitch | PLL | Polyphase | QAM | QDMA | Quantization | Quantizer | Radar | Random Noise | Reed Solomon | Remez | Resampling | RTDX | Sampling | Sharc | TI C6711 | Undersampling | Viterbi | Wavelets | White Noise | Wiener Filter | Windowing | XDS510PP | Z Transform

Sponsor

Industry's highest performing at the lowest power DSPs now as low as $5.00*
Start development today!
*volume pricing for 10ku

Discussion Groups

Free Online Books

See Also

Embedded SystemsFPGAElectronics

Discussion Groups | Comp.DSP | The Theory Behind Why DCT is Used for MFCC Feature Vectors?

There are 8 messages in this thread.

You are currently looking at messages 0 to 8.


The Theory Behind Why DCT is Used for MFCC Feature Vectors? - Raeldor - 2011-03-14 01:07:00

Hi All,

Does anyone have a good link to the theory behind why DCT is used for
MFCC feature vectors?  What exactly is it about the DCT values of the
mel frequency spectrum that makes it so appealing for voice
recognition.  I understand it makes the values less sensitive to
frequency, but how is that so?

Thanks
Rael
______________________________
New DSP Code Snippets Section now Live.   Learn more about the reward program for contributors here.

STUPIDENT::Re: The Theory Behind Why DCT is Used for MFCC Feature Vectors? - Vladimir Vassilevsky - 2011-03-14 10:21:00



Hey stupident. Everybody is already sick of your idiotic questions. Why 
don't you start with some ABC book on digital processing before 
venturing into the complicated matters of voice recognition?

Raeldor wrote:
> Hi All,
> 
> Does anyone have a good link to the theory behind why DCT is used for
> MFCC feature vectors?  What exactly is it about the DCT values of the
> mel frequency spectrum that makes it so appealing for voice
> recognition.  I understand it makes the values less sensitive to
> frequency, but how is that so?
> 
> Thanks
> Rael
______________________________
New DSP Code Snippets Section now Live.   Learn more about the reward program for contributors here.

Re: STUPIDENT::Re: The Theory Behind Why DCT is Used for MFCC Feature Vectors? - robert bristow-johnson - 2011-03-14 10:49:00

nice to see you back in form, Vlad.

sorta was wondering where you were during the DFT periodicity debate.

r b-j

On Mar 14, 10:21=A0am, Vladimir Vassilevsky <nos...@nowhere.com> wrote:
> Hey stupident. Everybody is already sick of your idiotic questions. Why
> don't you start with some ABC book on digital processing before
> venturing into the complicated matters of voice recognition?
>
> Raeldor wrote:
> > Hi All,
>
> > Does anyone have a good link to the theory behind why DCT is used for
> > MFCC feature vectors? =A0What exactly is it about the DCT values of the
> > mel frequency spectrum that makes it so appealing for voice
> > recognition. =A0I understand it makes the values less sensitive to
> > frequency, but how is that so?
>
> > Thanks
> > Rael
>
>

______________________________
New DSP Code Snippets Section now Live.   Learn more about the reward program for contributors here.

Re: STUPIDENT::Re: The Theory Behind Why DCT is Used for MFCC Feature Vectors? - Raeldor - 2011-03-15 01:17:00

On Mar 14, 7:21=A0am, Vladimir Vassilevsky <nos...@nowhere.com> wrote:
> Hey stupident. Everybody is already sick of your idiotic questions. Why
> don't you start with some ABC book on digital processing before
> venturing into the complicated matters of voice recognition?
>
>
>
>
>
>
>
> Raeldor wrote:
> > Hi All,
>
> > Does anyone have a good link to the theory behind why DCT is used for
> > MFCC feature vectors? =A0What exactly is it about the DCT values of the
> > mel frequency spectrum that makes it so appealing for voice
> > recognition. =A0I understand it makes the values less sensitive to
> > frequency, but how is that so?
>
> > Thanks
> > Rael

Feel like a big man now?
______________________________
New DSP Code Snippets Section now Live.   Learn more about the reward program for contributors here.

Re: The Theory Behind Why DCT is Used for MFCC Feature Vectors? - Raeldor - 2011-03-15 01:54:00

And I will solve the problem I'm trying to solve, with or without your
useless comments.  I always do.
______________________________
New DSP Code Snippets Section now Live.   Learn more about the reward program for contributors here.

STUDIOT::Re: The Theory Behind Why DCT is Used for MFCC Feature Vectors? - Vladimir Vassilevsky - 2011-03-15 10:13:00


Raeldor wrote:

> And I will solve the problem I'm trying to solve, with or without your
> useless comments.  I always do.

Sure. To start with, get down from the tree and try walking on the back 
paws.
______________________________
New DSP Code Snippets Section now Live.   Learn more about the reward program for contributors here.

Re: The Theory Behind Why DCT is Used for MFCC Feature Vectors? - Raeldor - 2011-03-15 11:08:00

On Mar 15, 7:13=A0am, Vladimir Vassilevsky <nos...@nowhere.com> wrote:
> Raeldor wrote:
> > And I will solve the problem I'm trying to solve, with or without your
> > useless comments. =A0I always do.
>
> Sure. To start with, get down from the tree and try walking on the back
> paws.

You know, I feel a little sorry for you.  It's obvious you've been
bullied or abused at some point and now feel like you need to prove
yourself by name-calling people who are taking an interest in a
subject you know something about.  It doesn't make you look any
smarter, it just makes you look like someone who needs help.  I've met
people like you before.  You might know a lot, and you might even be
smart, but you will never be content until you start treating people
with respect, because no matter how much you know or how clever you
are they'll never respect YOU until you do.
______________________________
New DSP Code Snippets Section now Live.   Learn more about the reward program for contributors here.

Re: The Theory Behind Why DCT is Used for MFCC Feature Vectors? - dbd - 2011-03-15 18:18:00

On Mar 13, 10:07=A0pm, Raeldor <rael...@gmail.com> wrote:
> Hi All,
>
> Does anyone have a good link to the theory behind why DCT is used for
> MFCC feature vectors? =A0What exactly is it about the DCT values of the
> mel frequency spectrum that makes it so appealing for voice
> recognition. =A0I understand it makes the values less sensitive to
> frequency, but how is that so?
>
> Thanks
> Rael

You might try googling on the title of the thread. After a couple of
pages of hits from this thread itself, there are a number of
potentially useful hits in the top 100. I think some of the papers by
Rosell may be responsive.

Dale B. Dalrymple
______________________________
New DSP Code Snippets Section now Live.   Learn more about the reward program for contributors here.