TY - GEN
T1 - Application of hidden markov chain and artificial neural networks in music recognition and classification
AU - Huang, Wei
AU - Zhang, Yuting
N1 - Publisher Copyright:
© 2020 ACM International Conference Proceeding Series. All rights reserved.
PY - 2020/1/4
Y1 - 2020/1/4
N2 - This paper presents a method for modeling and analyzing music recognition problems. The presented method is a hidden Markov chain (Hidden Markov Models, HMM) combined with artificial neural networks (Artificial Neural Network, ANN). This method can be considered as an improvement of the classical HMM method with the use of the ANN. In this paper, the modeling of music recognition and algorithm implementation is discussed. The algorithm is then applied to the actual music classification problem. An analysis of its performance as compared to the conventional HMM-only method and ANN-only method is provided. The hidden Markov process model is faster but suffers severe performance degradation when dealing with a large number of categories. Meanwhile, the artificial neural network method has better classification performance, but the computation complexity is higher compared to the hidden Markov model. The proposed method, which is the combination of the two methods, can improve the recognition accuracy of the HMM by 4 to 5 percent while ensuring a similar time-complexity as the conventional HMM.
AB - This paper presents a method for modeling and analyzing music recognition problems. The presented method is a hidden Markov chain (Hidden Markov Models, HMM) combined with artificial neural networks (Artificial Neural Network, ANN). This method can be considered as an improvement of the classical HMM method with the use of the ANN. In this paper, the modeling of music recognition and algorithm implementation is discussed. The algorithm is then applied to the actual music classification problem. An analysis of its performance as compared to the conventional HMM-only method and ANN-only method is provided. The hidden Markov process model is faster but suffers severe performance degradation when dealing with a large number of categories. Meanwhile, the artificial neural network method has better classification performance, but the computation complexity is higher compared to the hidden Markov model. The proposed method, which is the combination of the two methods, can improve the recognition accuracy of the HMM by 4 to 5 percent while ensuring a similar time-complexity as the conventional HMM.
KW - Artificial neural network
KW - Hidden Markov models
KW - Machine learning
KW - Music classification
UR - http://www.scopus.com/inward/record.url?scp=85082168882&partnerID=8YFLogxK
U2 - 10.1145/3379247.3379276
DO - 10.1145/3379247.3379276
M3 - Conference contribution
AN - SCOPUS:85082168882
T3 - ACM International Conference Proceeding Series
SP - 49
EP - 53
BT - ICCDE 2020 - 2020 the 6th International Conference on Computing and Data Engineering; AIEE 2020 - 2020 International Conference on Artificial Intelligence in Electronics Engineering
PB - Association for Computing Machinery
T2 - 6th International Conference on Computing and Data Engineering, ICCDE 2020 and 2020 International Conference on Artificial Intelligence in Electronics Engineering, AIEE 2020
Y2 - 4 January 2020 through 6 January 2020
ER -