Music Information Retrieval

Music Information Retrieval (MIR) focuses on developing computational methods to analyze, organize, and retrieve information from music. Current research emphasizes improving automatic music transcription (using convolutional recurrent neural networks and transformers), developing robust genre classification models (often leveraging deep learning on specialized datasets), and creating explainable AI for tasks like difficulty estimation. These advancements are significant for music education, enhancing music discovery and recommendation systems, and fostering more effective human-computer musical collaboration.