PBSCR: The Piano Bootleg Score Composer Recognition Dataset
This dataset addresses the need for standardized, large-scale data in music information retrieval, specifically for composer recognition, though it is incremental as it builds on existing bootleg score representations and IMSLP resources.
The authors introduced the PBSCR dataset for composer recognition of classical piano music, providing 40,000 labeled images for a 9-class task and 100,000 for a 100-class task, along with 29,310 unlabeled images for pretraining, to enable large-scale research with modern architectures.
This article motivates, describes, and presents the PBSCR dataset for studying composer recognition of classical piano music. Our goal was to design a dataset that facilitates large-scale research on composer recognition that is suitable for modern architectures and training practices. To achieve this goal, we utilize the abundance of sheet music images and rich metadata on IMSLP, use a previously proposed feature representation called a bootleg score to encode the location of noteheads relative to staff lines, and present the data in an extremely simple format (2D binary images) to encourage rapid exploration and iteration. The dataset itself contains 40,000 62x64 bootleg score images for a 9-class recognition task, 100,000 62x64 bootleg score images for a 100-class recognition task, and 29,310 unlabeled variable-length bootleg score images for pretraining. The labeled data is presented in a form that mirrors MNIST images, in order to make it extremely easy to visualize, manipulate, and train models in an efficient manner. We include relevant information to connect each bootleg score image with its underlying raw sheet music image, and we scrape, organize, and compile metadata from IMSLP on all piano works to facilitate multimodal research and allow for convenient linking to other datasets. We release baseline results in a supervised and low-shot setting for future works to compare against, and we discuss open research questions that the PBSCR data is especially well suited to facilitate research on.