It is possible to download software here.
Publication to be cited
MEKYSKA, J.; FAÚNDEZ ZANUY, M.; SMÉKAL, Z.; FABREGAS, J. Score Fusion in Text- Dependent Speaker Recognition Systems. Lecture Notes in Computer Science, 2011, vol. 6800, no. 12, pp. 120-132. ISSN: 0302-9743.
This software recognizes the isolate digits. It is possible to recognize digits 0-9. Particular digits’ templates are stored in folder “wav”. It is possible to change or extend these templates which is recommended especially in case the user is about to change a microphone (there can be problems with a channel distortion). To recognize the digit and to calculate a distance between template and unknown signal there was used algorithm of dynamic time warping. This software can be used to develop or test the classifiers, speech features or it can be used for education in the field of pattern recognition.
This software uses Dynamic Time Warp (DTW) toolbox that can be downloaded under link http://labrosa.ee.columbia.edu/matlab/dtw/ The software will not work properly without this software. Please before use, download this toolbox end extract its functions directly to a folder where file digits_recognition.m is stored. The software must be run in a 32-bit version of MATLAB (Data Acquisition Toolbox DAQ must be installed). Before you run the software please ensure yourself that the microphone is connected to a computer. Acoustic signal from microphone, connected after a start of software, will not be processed. The software can be run using command digits_recognition Set the length of segment, press button Start and utter a digit. The software will then automatically stop the acquisition and displays the recognized digit. Beside the length of speech segment it is possible to change also the sampling frequency.
This work was supported by projects VG20102014033 and FEKT-S-11-17. The described research was performed in laboratories supported by the SIX project; the registration number CZ.1.05/2.1.00/03.0072, the operational program Research and Development for Innovation.