Music performance assessment is a highly subjective task often relying on experts to gauge both the technical and aesthetic aspects of the performance from the audio signal. This article explores the task of building computational models for music performance assessment, i.e., analyzing an audio recording of a performance and rating it along several criteria such as musicality, note accuracy, etc. Much of the earlier work in this area has been centered around using hand-crafted features intended to capture relevant aspects of a performance. However, such features are based on our limited understanding of music perception and may not be optimal. In this article, we propose using Deep Neural Networks (DNNs) for the task and compare their performance against a baseline model using standard and hand-crafted features. We show that, using input representations at different levels of abstraction, DNNs can outperform the baseline models across all assessment criteria. In addition, we use model analysis techniques to further explain the model predictions in an attempt to gain useful insights into the assessment process. The results demonstrate the potential of using supervised feature learning techniques to better characterize music performances.

Deep generative models for symbolic music are typically designed to model temporal dependencies in music so as to predict the next musical event given previous events. In many cases, such models are expected to learn abstract concepts such as harmony, meter, and rhythm from raw musical data without any additional information. In this study, we investigate the effects of explicitly conditioning deep generative models with musically relevant information. Specifically, we study the effects of four different conditioning inputs on the performance of a recurrent monophonic melody generation model. Several combinations of these conditioning inputs are used to train different model variants which are then evaluated using three objective evaluation paradigms across two genres of music. The results indicate musically relevant conditioning significantly improves learning and performance, and reveal how this information affects learning of musical features related to pitch and rhythm. An informal subjective evaluation suggests a corresponding improvement in the aesthetic quality of generations.

Selective manipulation of data attributes using deep generative models is an active area of research. In this paper, we present a novel method to structure the latent space of a Variational Auto-Encoder (VAE) to encode different continuous-valued attributes explicitly. This is accomplished by using an attribute regularization loss which enforces a monotonic relationship between the attribute values and the latent code of the dimension along which the attribute is to be encoded. Consequently, post-training, the model can be used to manipulate the attribute by simply changing the latent code of the corresponding regularized dimension. The results obtained from several quantitative and qualitative experiments show that the proposed method leads to disentangled and interpretable latent spaces that can be used to effectively manipulate a wide range of data attributes spanning image and symbolic music domains.