Counterpoint by Convolution
This work addresses the challenge of creating more realistic and flexible music generation models for AI and music composition applications, though it is incremental as it builds on existing orderless NADE and Gibbs sampling techniques.
The paper tackled the problem of generating music by moving away from chronological composition models to better mimic human composers' nonlinear process, using a convolutional neural network and blocked Gibbs sampling to complete partial scores, which improved sample quality over ancestral sampling as shown by log-likelihood and human evaluation.
Machine learning models of music typically break up the task of composition into a chronological process, composing a piece of music in a single pass from beginning to end. On the contrary, human composers write music in a nonlinear fashion, scribbling motifs here and there, often revisiting choices previously made. In order to better approximate this process, we train a convolutional neural network to complete partial musical scores, and explore the use of blocked Gibbs sampling as an analogue to rewriting. Neither the model nor the generative procedure are tied to a particular causal direction of composition. Our model is an instance of orderless NADE (Uria et al., 2014), which allows more direct ancestral sampling. However, we find that Gibbs sampling greatly improves sample quality, which we demonstrate to be due to some conditional distributions being poorly modeled. Moreover, we show that even the cheap approximate blocked Gibbs procedure from Yao et al. (2014) yields better samples than ancestral sampling, based on both log-likelihood and human evaluation.