Genre classification of songs using neural network

Abstract

The objective here is to eliminate the manual work of classifying genres of song in each song. With this startup work songs can be classified in real-time and proposed parallel architecture can be implemented on the multi-processing system as well. In this paper a set of features are obtained like beats/tempo, energy, loudness, speechiness, valence, danceability, acousticness, discrete wavelet transform etc., using Echonest libraries and are fed into the Parallel Multi-Layer Perceptron Network to obtain the genres of the song. The proposed scheme has an accuracy of 85% when used to classify two genres of songs that are Sufi and Classical.

Identifying Mood in Music Using Deep Learning

“…Many research papers are dedicated to signal processing acoustic feature. Acoustic feature can be classified as [5] rhythmic features [3] like beat, loudness, DWT and timbral texture feature [2] like MFCC that are used to discriminate music [6], pitch feature [7] that deals with detection of the fear emotion.…”

“…In a study by Asuman et al [3], classification is achieved using multilayer perceptron network for the songs taxonomy by genres. You et al presented a method to classify music in an effective way based on vocalists [6].…”

“…In literature, feature extraction in machine learning is performed as the preprocessing task [2][3][4]. Certain features like mel-frequency cepstral coefficients (MFCC) work on timbral information, and signal processing acoustic feature works on rhythmic and pitch information.…”

Music Genre Classification using Spectral Analysis Techniques With Hybrid Convolution-Recurrent Neural Network

“…When apply Recurrent Neural Network, they achieved highest accuracy among the others with 86%. Anshuman Goel et al [6], they used 8 feature extraction from the audio file, that is, beat periodicity, loudness, energy, speechness, acousticness, valence, danceability, discrete wavelet transform (DWT). Then these features are fed to the neural network.…”

MFCC, Spectral and Temporal Feature based Emotion Identification in Songs

“…This toolbox has also been used in various works such as [7] and [8] and known to be very useful in classification problems for audio samples. Some preprocessing steps such as min-max normalization were also performed before the proceeding to the training phase.…”