Question

• Record your voice for a few seconds with fs=16000 • Divide the signal into frames of 25 ms with 10 ms overlap. Apply Hamming window to the frames, then compute the power spectrum (abs(fft(frame,512)).^2). • Design a filterbank with number of filters equal to the last two digits of your student I.D. + 15 • Use the formula on page 30 if your student ID is an even number, else, use the one on page 31 • Multiply the power spectrum with each of the filters. (Bonus points if you can do this part with a single operation without any loops, conditions, etc.) • Take the log of the result (should be a matrix with one dimension equal to the number of filters and the other one equal to number of frames) • Apply discrete cosine transform (dct function of Matlab), and chose the first 12 coefficients, delete the others. • Bonus points if you calculate delta coefs. using the

          • Record your voice for a few seconds with fs=16000
• Divide the signal into frames of 25 ms with 10 ms overlap.
Apply Hamming window to the frames, then compute the
power spectrum (abs(fft(frame,512)).^2).
• Design a filterbank with number of filters equal to the last
two digits of your student I.D. + 15
• Use the formula on page 30 if your student ID is an even
number, else, use the one on page 31
• Multiply the power spectrum with each of the filters.
(Bonus points if you can do this part with a single
operation without any loops, conditions, etc.)
• Take the log of the result (should be a matrix with one
dimension equal to the number of filters and the other
one equal to number of frames)
• Apply discrete cosine transform (dct function of
Matlab), and chose the first 12 coefficients, delete the
others.
• Bonus points if you calculate delta coefs. using the

• Record your voice for a few seconds with fs=16000
• Divide the signal into frames of 25 ms with 10 ms overlap.
Apply Hamming window to the frames, then compute the
power spectrum (abs(fft(frame,512)).^2).
• Design a filterbank with number of filters equal to the last
two digits of your student I.D. + 15
• Use the formula on page 30 if your student ID is an even
number, else, use the one on page 31
• Multiply the power spectrum with each of the filters.
(Bonus points if you can do this part with a single
operation without any loops, conditions, etc.)
• Take the log of the result (should be a matrix with one
dimension equal to the number of filters and the other
one equal to number of frames)
• Apply discrete cosine transform (dct function of
Matlab), and chose the first 12 coefficients, delete the
others.
• Bonus points if you calculate delta coefs. using the

Added by Garrett R.

Question

Please give Ace some feedback