Modeling Musical Onset Probabilities via Neural Distribution Learning

2020/02/10 by Huh, Jaesung, Martinsson, Egil, Kim, Adrian +1
#Audio and Speech Processing (eess.AS) #FOS: Computer and information sciences #FOS: Electrical engineering #Machine Learning (cs.LG) #Sound (cs.SD) #electronic engineering #information engineering

paper · doi:10.48550/arxiv.2002.03559

Abstract

Musical onset detection can be formulated as a time-to-event (TTE) or time-since-event (TSE) prediction task by defining music as a sequence of onset events. Here we propose a novel method to model the probability of onsets by introducing a sequential density prediction model. The proposed model estimates TTE & TSE distributions from mel-spectrograms using convolutional neural networks (CNNs) as a density predictor. We evaluate our model on the Bock dataset show-ing comparable results to previous deep-learning models.

Related

• Towards an efficient deep learning model for musical onset detection
• Striking a New Chord: Neural Networks in Music Information Dynamics
• Mind the beat: detecting audio onsets from EEG recordings of music listening
• Estimating Musical Surprisal in Audio
• An Interactive Musical Prediction System with Mixture Density Recurrent Neural Networks