Tempo estimation via self-supervised learning

• Ano de defesa: 2023
• Autor(a) principal: Morais, Giovana Vieira de
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Biblioteca Digitais de Teses e Dissertações da USP
• Palavras-chave em Português: Estimação de andamento; Self-supervised learning; Tempo estimation
• Link de acesso: https://www.teses.usp.br/teses/disponiveis/45/45134/tde-01042024-142309/

Self-supervision methods learn representations by solving pretext tasks that do not require human-generated labels, alleviating the need for time-consuming annotations. These methods have been applied in computer vision, natural language processing, environmental sound analysis, and recently in music information retrieval, e.g. for pitch estimation. Particularly in the context of music, there are few insights about the fragility of these models regarding different distributions of data, and how they could be mitigated. In this work, we explore these questions by analyzing a self-supervised model for pitch estimation that we adapted for tempo estimation via rigorous experimentation with synthetic data and contrasting its behavior with real data. We discuss the design choices regarding the method and the input representation. Finally, we study the relationship between the input representation and data distribution for self-supervised tempo estimation.