Towards the digital music library: tune retrieval from acoustic input
Proceedings of the first ACM international conference on Digital libraries
Melodic matching techniques for large music databases
MULTIMEDIA '99 Proceedings of the seventh ACM international conference on Multimedia (Part 1)
HMM-based musical query retrieval
Proceedings of the 2nd ACM/IEEE-CS joint conference on Digital libraries
A Survey of Longest Common Subsequence Algorithms
SPIRE '00 Proceedings of the Seventh International Symposium on String Processing Information Retrieval (SPIRE'00)
Warping indexes with envelope transforms for query by humming
Proceedings of the 2003 ACM SIGMOD international conference on Management of data
Algorithmic Clustering of Music Based on String Compression
Computer Music Journal
Anchor space for classification and similarity measurement of music
ICME '03 Proceedings of the 2003 International Conference on Multimedia and Expo - Volume 2
Aggregate features and ADABOOST for music classification
Machine Learning
Music genre classification using MIDI and audio features
EURASIP Journal on Applied Signal Processing
A comprehensive trainable error model for sung music queries
Journal of Artificial Intelligence Research
Challenging Uncertainty in Query by Humming Systems: A Fingerprinting Approach
IEEE Transactions on Audio, Speech, and Language Processing
A Survey of Audio-Based Music Classification and Annotation
IEEE Transactions on Multimedia
Hum-a-song: a subsequence matching with gaps-range-tolerances query-by-humming system
Proceedings of the VLDB Endowment
A survey of query-by-humming similarity methods
Proceedings of the 5th International Conference on PErvasive Technologies Related to Assistive Environments
Hi-index | 0.00 |
Automatic music genre classification is a task that has attracted the interest of the music community for more than two decades. Music can be of high importance within the area of assistive technologies as it can be seen as an assistive technology with high therapeutic and educational functionality for children and adults with disabilities. Several similarity methods and machine learning techniques have been applied in the literature to deal with music genre classification, and as a result data mining and Music Information Retrieval (MIR) are strongly interconnected. In this paper, we deal with music genre classification for symbolic music, and specifically MIDI, by combining the recently proposed novel similarity measure for sequences, SMBGT, with the k-Nearest Neighbor (k-NN) classifier. For all MIDI songs we first extract all of their channels and then transform each channel into a sequence of 2D points, providing information for pitch and duration of their music notes. The similarity between two songs is found by computing the SMBGT for all pairs of the songs' channels and getting the maximum pairwise channel score as their similarity. Each song is treated as a query to which k-NN is applied, and the returned genre of the classifier is the one with the majority of votes in the k neighbors. Classification accuracy results indicate that there is room for improvement, especially due to the ambiguous definitions of music genres that make it hard to clearly discriminate them. Using this framework can also help us analyze and understand potential disadvantages of SMBGT, and thus identify how it can be improved when used for classification of real-time sequences.