Musical instrument recognition based on class pairwise feature selection

In this work, musical instrument recognition is considered on solo music from real world performance. A large sound database is used that consists of musical phrases excerpted from commercial recordings with different instrument instances, different players, and varying recording conditions. The proposed recognition scheme exploits class pairwise feature selection based on inertia ratio maximization. Moreover, new signal processing features based on octave band energy measures are introduced that prove to be useful. Classification is performed using Gaussian Mixture Models in a one vs one fashion in association with a data rescaling procedure as pre-processing. Experimental results show that substantial improvement in recognition success is thus achieved.

• Type: Conference
• Year: 2004
• Where: ISMIR
• Authors: Slim Essid, Gaël Richard, Bertrand David
• Keywords: Inertia Ratio Maximization, ISMIR 2004, Musical Instrument Recognition, Scheme Exploits Class