Technologies for the automatic description of music scores from varied symbolic encodings, addressing e.g. summarisation, pattern finding and identification, similarity, analysis of melody, rhythm or tonality. Present limitations include a lack of accurate models for capturing mid- and high-level concepts in meaningful ways. Musicological analysis requires addressing the specificities of the repertoire and research questions, often including novel notation or encodings. Finally, of utmost importance is the quality of the encodings.
TROMPA will advance the state of the art by enabling higher quality encodings supported by the crowd, combining symbolic and multimodal descriptions, and designing meaningful descriptors for different types of users in the pilots and according to the repertoire and the research questions that may arise. Where possible, existing standards (such as MEI) will be adopted or repurposed and/or extended to ensure sustainable outcomes.