Paper Title
Multi-Instrument Detection in Culture Musics Using Machine Learning Models

Abstract - With music universalisation, some cultural instruments have gradually become familiar to the public. There is a need for an application that helps people recognize the musical instrument. Based on this premise, this paper proposes an instrument detection system to support music recognition of non-mainstream musical instruments. This research is based on many rare musical instrument audio signal samples collected by Cultural Infusion. On this basis, the audio signal is analysed to obtain characteristic parameters of different instruments, which are integrated into a database. The instrument detection system is generated by training the data through different Machine Learning models,and the performance will be discussed. Keywords - Polyphony, Music Information Retrieval, Multi-class Classification, Deep Learning, Attention Mechanism