A study on the automatic classification of musical instrument sounds ispresented. For this purpose a large database of musical instrument soundswas built, which consists of both solo and duet stereo recordings. Theclassification process of musical instrument sounds is done on the basis ofsome soft computing techniques, such as neural networks. The results ofthe classification are given as a percentage of musical instrument soundsproperly recognized by the system. A discussion of the system efficiencyand of its limitations is presented. Conclusions and remarks concerningfurther development of this study are included.
展开▼