Musical Chord Recognition based on Robust Pitch Class Profile and Metric Learning Support Vector Machine

Chord recognition is an important aspect of Music Information Retrieval, which plays an important role in information processing, musical structure analysis and recommender system. In order to reduce the influence of voice on chord progression and recovery harmonic information of chord, paper modeled harmonic and voice component and constructed a two-target optimal problem. Solving the optimal problem, paper reconstructed harmonic structure and removed sparse voice. Then, through performing a pitch mapping step, paper obtained robust pitch class profile. At last, obtaining a Mahalanobis matrix from feature space as metric matrix, paper replaced Euclidean Metric utilized in Radius Basis Function of SVM with Mahalanobis Metric. Mahalanobis Metric contain distribution information of specified real dataset, so the classification result is more robust. Compared with currently popular chord estimation algorithm, Results show proposed system improves the accuracy ratio of 3.5%~12.2% on chord recognition.