A Dataset for Ethiopian Begena songs

This dataset was meticulously crafted during the dataset collection process for the MSC thesis titled "Integrating Music Genre Classification for Precision in Constructing an Automatic Music Transcription Model for Ethiopian Begena" by Bisrat Getnet, advised by Tesfa Tegegne (Ph.D.). The collaborative efforts of a dedicated team, including Ashenafi Tezera and Nathnael Yilkal, contributed to the successful development of this dataset. While Bisrat and Ashenafi played roles in instrument performance, Nathnael took charge of recording and generating the accompanying MIDI files. Main Objectives: Automatic Music Transcription: The dataset was initially created to support the Automatic Music Transcription aspect of the thesis. Music Genre Classification: Subsequently, it has been harnessed for the Music Genre Classification aspect of the thesis. Music Generation and Other MIR Researches: Beyond its primary objectives, the dataset is versatile and can be applied to explore Music Generation and other Music Information Retrieval (MIR) research endeavors. Contents: The zip file comprises three folders, each representing a primary scale in Ethiopian music: Cherenet, Wanen, and Selameta. Within each scale folder, two subfolders named MIDI and FLAC organize the dataset. File Types: The FLAC folder contains .flac files, representing the audio recordings. The MIDI folder contains .mid files encapsulating note events extracted from the corresponding FLAC files. Scale Categories: The three main scales (Cherenet, Wanen, and Selameta) form the fundamental basis for exploring music scale classification within the context of Ethiopian music. This dataset is purposefully curated to facilitate research and exploration in Music Genre Classification, Music Scale Classification, Automatic Music Transcription, Music Generation, and various other Music Information Retrieval research areas. Its unique focus on Ethiopian music scales contributes to its significance in diverse applications within the field.