Shipei Liu

Guowei Wu, Shipei Liu, Xiaoya Fan

Symbolic Music Generation relies on the contextual representation capabilities of the generative model, where the most prevalent approach is the Transformer-based model. The learning of musical context is also related to the structural elements in music, i.e. intro, verse, and chorus, which are currently overlooked by the research community. In this paper, we propose a hierarchical Transformer model to learn multi-scale contexts in music. In the encoding phase, we first designed a Fragment Scope Localization layer to syncopate the music into chords and sections. Then, we use a multi-scale attention mechanism to learn note-, chord-, and section-level contexts. In the decoding phase, we proposed a hierarchical Transformer model that uses fine-decoders to generate sections in parallel and a coarse-decoder to decode the combined music. We also designed a Music Style Normalization layer to achieve a consistent music style between the generated sections. Our model is evaluated on two open MIDI datasets, and experiments show that our model outperforms the best contemporary music generative models. More excitingly, the visual evaluation shows that our model is superior in melody reuse, resulting in more realistic music.

Shipei Liu, Xiaoya Fan, Guowei Wu

Existing music generation models are mostly language-based, neglecting the frequency continuity property of notes, resulting in inadequate fitting of rare or never-used notes and thus reducing the diversity of generated samples. We argue that the distribution of notes can be modeled by translational invariance and periodicity, especially using diffusion models to generalize notes by injecting frequency-domain Gaussian noise. However, due to the low-density nature of music symbols, estimating the distribution of notes latent in the high-density solution space poses significant challenges. To address this problem, we introduce the Music-Diff architecture, which fits a joint distribution of notes and accompanying semantic information to generate symbolic music conditionally. We first enhance the fragmentation module for extracting semantics by using event-based notations and the structural similarity index, thereby preventing boundary blurring. As a prerequisite for multivariate perturbation, we introduce a joint pre-training method to construct the progressions between notes and musical semantics while avoiding direct modeling of low-density notes. Finally, we recover the perturbed notes by a multi-branch denoiser that fits multiple noise objectives via Pareto optimization. Our experiments suggest that in contrast to language models, joint probability diffusion models perturbing at both note and semantic levels can provide more sample diversity and compositional regularity. The case study highlights the rhythmic advantages of our model over language- and DDPMs-based models by analyzing the hierarchical structure expressed in the self-similarity metrics.