About

In recent years, there has been a notable and substantial increase in the widespread adoption and popularity of streaming music services. This has drawn out a range of reactions, including both optimism and apprehension, regarding their effects on revenue generated from recorded music. As a result, there is an increasing need for efficient and accurate methods to classify music into genres. The rapid evolution of advanced multimedia technologies has made abundant musical resources accessible online, generating a continuous interest in categorizing diverse music genres.

This research is dedicated to develop more accurate and robust classification of music genres. With the continuous growth of digital music libraries and the pressing demand for streamlined genre labeling, this study aims to substantially enhance the accuracy and efficiency of music genre identification using Ensemble Learning.

Furthermore, Advocates will impose limitations on permissible music genres, stipulating that only Blues, Classical, Country, Disco, Hip-hop, Jazz, Metal, Pop, Reggae, and Rock are permitted. Any other genres (e.g., datasets) will be excluded to minimize problems when interpreting the results. In terms of datasets, Regarding datasets, the data will be sourced from GTZAN, a widely recognized dataset utilized in music genre classification research. This dataset includes ten genres, each with 100 audio files and another 900 audio files that are not related to GTZAN, which are from different songs. Every audio file is 30 seconds long. Ultimately, the suggested model is anticipated to yield a more precise, quicker, and more effective result than the study that solely used convolutional neural networks. The study exclusively utilizes WAV file format due to its compatibility with the selected machine learning algorithms. Other formats, such as MP3, are not currently supported.