Selected Research interests:
Application of pattern classification techniques to the classification of audio frequency waveforms, eg. musical instrument sounds. These techniques range from statistical, neural as well as non-parametric techniques, such as NNCs. Unsupervised learning or clustering techniques are also being investigated
Selected Publications:
Kaminskyj, I., "Multidimensional Scaling Analysis of Musical Instrument Sounds' Spectra" Proc. Aust. Comp. Music Assoc. Conf. 1999 Wellington, New Zealand, July 1999, pp. 36-39
Kaminskyj, I., "Automatic Musical Vibrato Detection System" Proc. Aust. Comp. Music Assoc. Conf. 1998 Canberra, Australia, July 1998, pp. 9-15
Kaminskyj, I., and Voumard, P., "Enhanced Automatic Source Identification of Monophonic Musical Instrument Sounds", 1996 Aust. NZ. Conf.
Intelligent Information Systems, Adelaide, SA, Nov. 1996, pp. 76-79
Kaminskyj, I., and Materka, A., "Automatic Source Identification of Monophonic Musical Instrument Sounds", IEEE Int. Conf. Neural Networks
1995, Vol. 1, Perth, WA 1995, pp. 189-194.
TOP
