Locally Activated Gated Neural Network for Automatic Music Genre Classification

Article Computer Science, Artificial Intelligence

A hybrid deep learning approach for classification of music genres using wavelet and spectrogram analysis

Kalyan Kumar Jena et al.

NEURAL COMPUTING & APPLICATIONS (2023)

Article Chemistry, Multidisciplinary

A Hybrid CNN and RNN Variant Model for Music Classification

Mohsin Ashraf et al.

Summary: Music genre classification plays a significant role in organizing large collections of music for information retrieval. Traditional methods using handcrafted features struggle to accurately determine the original music characteristics. This paper proposes a hybrid architecture of CNN and RNN variants, such as LSTM, Bi-LSTM, GRU, and Bi-GRU, to address the limitations of existing neural network classification models. Experimental results show that the CNN-Bi-GRU architecture using Mel-spectrogram achieves the highest accuracy of 89.30%, while the CNN-LSTM hybridization using MFCC achieves the highest accuracy of 76.40%.

APPLIED SCIENCES-BASEL (2023)

Add to Collection

Article Mathematics

Machine Learning for Music Genre Classification Using Visual Mel Spectrum

Yu-Huei Cheng et al.

Summary: Music is a convenient and easy-to-use tool for releasing stress, and listening to appropriate music can help alleviate stress. To effectively search for specific genres of music, researchers propose a novel method based on the visual Mel spectrum for music genre classification, using YOLOv4 as the neural network architecture.

MATHEMATICS (2022)

Add to Collection

Article Computer Science, Information Systems

Large-Scale Music Genre Analysis and Classification Using Machine Learning with Apache Spark

Mousumi Chaudhury et al.

Summary: This research paper analyzes the use of machine learning models supported by Apache Spark to classify music genres in an online music library. The experimental results demonstrate that the random forest classifier outperforms other classifiers, achieving 90% accuracy in music genre classification.

ELECTRONICS (2022)

Add to Collection

Proceedings Paper Acoustics

S3T: SELF-SUPERVISED PRE-TRAINING WITH SWIN TRANSFORMER FOR MUSIC CLASSIFICATION

Hang Zhao et al.

Summary: In this paper, a self-supervised pre-training method called S3T is proposed, which utilizes Swin Transformer to learn meaningful music representations from easily accessible unlabeled music data. Experimental results show that S3T outperforms previous methods in music classification and achieves improvements in label efficiency.

2022 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP) (2022)

Add to Collection

Article Computer Science, Information Systems

Jazz Bass Transcription Using a U-Net Architecture

Jakob Abesser et al.

Summary: In this paper, the U-net deep neural network architecture is adapted for bass transcription, with pitch shifting and random equalization investigated as data augmentation techniques. The study examines the importance of skip connections between encoder and decoder layers, data augmentation strategies, and overall model capacity on system performance. Results show that using skip connections with training and validation sets can improve bass transcription performance.

ELECTRONICS (2021)

Add to Collection

Article Computer Science, Information Systems

A Middle-Level Learning Feature Interaction Method with Deep Learning for Multi-Feature Music Genre Classification

Jinliang Liu et al.

Summary: Music genre classification has become an interesting research area, with the multi-feature model considered as an ideal technology for classification. In this study, a method based on deep learning that focuses on intermediate-level feature interaction is proposed, which significantly improves the accuracy of music genre classification. Experimental results show that this method outperforms most current methods, with a classification accuracy of up to 93.65% on the GTZAN dataset.

ELECTRONICS (2021)

Add to Collection

Article Engineering, Electrical & Electronic