welee666 / audioprocessing-toolbox

extract the time domain or frequent domain features from wav format audio

AudioProcessing

An toolbox for audio processing in Matlab,the verison of Matlab is R2016a.
The audio sample in this toobox is 8000Hz, 16 bits.

1. getAudioFingerpriner

• getAudioFingerprinter calcuates the difference of the energy of each frequency sub-band as shown below.
  

• Specifically, the extraction process including the following steps:

1. divide audio into frames, the overlap length is alterable
2. do fft for each frame
3. divide frequency spectrum into 32 frequency sub bands
4. calcuate the energy of each sub band
5. calcuate audio fingerprinter as follow
   

• The result is shown below

2. getZCR

• getZCR calcuates the zero-crossing rate which can be used to predict the frequency roughly.
• Specifically, the extraction process including the following steps:

1. divide audio into frames
2. calcuate zcr of each frame using the following equation
   
   where sgn(x) is given by
   

• The result is shown below
  

3. getLandmarks

• getLandmarks calcuates the max-enegry points(also called landmark) in the frequency spectrum.

• Specifically, the extraction process including the following steps:

1. divide audio into frames, the overlap length is alterable
2. do fft for each frame
3. calcuate the energy of frequency
4. calcuate the local maximum as landmark

• The result is shown below
  

4. getEnergy

• getEnergy calcuates the short-time energy.

• Specifically, the extraction process including the following steps:
  1.divide audio into frames, the overlap length is alterable
  2.calcuate the energy using the following equation:
  

• The result is shown below
  

5. getSpecgram

• getSpecgram calcuates the spectgram.
• Specifically, the extraction process including the following steps:

1. divide signal into frames in time domain
2. do fft of each frame

• The result is shown below
  

6. getLMSPredictor

• getLMSPredictor trains least mean square(LMS) adaptive predictor.