Models for audio generation are typically trained on hours of recordings. Here, we illustrate that capturing the essence of an audio source is typically possible from as little as a few tens of seconds from a single training signal. Specifically, we present a GAN-based generative model that can be trained on one short audio signal from any domain (e.g. speech, music, etc.) and does not require pre-training or any other form of external supervision. Once trained, our model can generate random samples of arbitrary duration that maintain semantic similarity to the training waveform, yet exhibit new compositions of its audio primitives. This enables a long line of interesting applications, including generating new jazz improvisations or new a-cappella rap variants based on a single short example, producing coherent modifications to famous songs (e.g. adding a new verse to a Beatles song based solely on the original recording), filling-in of missing parts (inpainting), extending the bandwidth of a speech signal (super-resolution), and enhancing old recordings without access to any clean training example. We show that in all cases, no more than 20 seconds of training audio commonly suffice for our model to achieve state-of-the-art results. This is despite its complete lack of prior knowledge about the nature of audio signals in general.
In this paper, we propose a generic technique to model temporal dependencies and sequences using a combination of a recurrent neural network and a Deep Belief Network. Our technique, RNN-DBN, is an amalgamation of the memory state of the RNN that allows it to provide temporal information and a multi-layer DBN that helps in high level representation of the data. This makes RNN-DBNs ideal for sequence generation. Further, the use of a DBN in conjunction with the RNN makes this model capable of significantly more complex data representation than an RBM. We apply this technique to the task of polyphonic music generation.
Music generation has generally been focused on either creating scores or interpreting them. We discuss differences between these two problems and propose that, in fact, it may be valuable to work in the space of direct $\it performance$ generation: jointly predicting the notes $\it and$ $\it also$ their expressive timing and dynamics. We consider the significance and qualities of the data set needed for this. Having identified both a problem domain and characteristics of an appropriate data set, we show an LSTM-based recurrent network model that subjectively performs quite well on this task. Critically, we provide generated examples. We also include feedback from professional composers and musicians about some of these examples.
This article discusses the extension of the notion of context from linguistics to the domain of music. In language, the statistical regularity known as Zipf's law -which concerns the frequency of usage of different words- has been quantitatively related to the process of text generation. This connection is established by Simon's model, on the basis of a few assumptions regarding the accompanying creation of context. Here, it is shown that the statistics of note usage in musical compositions are compatible with the predictions of Simon's model. This result, which gives objective support to the conceptual likeness of context in language and music, is obtained through automatic analysis of the digital versions of several compositions. As a by-product, a quantitative measure of context definiteness is introduced and used to compare tonal and atonal works.
Today, a massive amount of musical knowledge is stored in written form, with testimonies dated as far back as several centuries ago. In this work, we present different Natural Language Processing (NLP) approaches to harness the potential of these text collections for automatic music knowledge discovery, covering different phases in a prototypical NLP pipeline, namely corpus compilation, text-mining, information extraction, knowledge graph generation and sentiment analysis. Each of these approaches is presented alongside different use cases (i.e., flamenco, Renaissance and popular music) where large collections of documents are processed, and conclusions stemming from data-driven analyses are presented and discussed.
Current music recommender systems typically act in a greedy fashion by recommending songs with the highest user ratings. Greedy recommendation, however, is suboptimal over the long term: it does not actively gather information on user preferences and fails to recommend novel songs that are potentially interesting. A successful recommender system must balance the needs to explore user preferences and to exploit this information for recommendation. This paper presents a new approach to music recommendation by formulating this exploration-exploitation trade-off as a reinforcement learning task called the multi-armed bandit. To learn user preferences, it uses a Bayesian model, which accounts for both audio content and the novelty of recommendations. A piecewise-linear approximation to the model and a variational inference algorithm are employed to speed up Bayesian inference. One additional benefit of our approach is a single unified model for both music recommendation and playlist generation. Both simulation results and a user study indicate strong potential for the new approach.
With increasing amounts of music being digitally transferred from production to distribution, automatic means of determining media quality are needed. Protection mechanisms in digital audio processing tools have not eliminated the need of production entities located downstream the distribution chain to assess audio quality and detect defects inserted further upstream. Such analysis often relies on the received audio and scarce meta-data alone. Deliberate use of artefacts such as clicks in popular music as well as more recent defects stemming from corruption in modern audio encodings call for data-centric and context sensitive solutions for detection. We present a convolutional network architecture following end-to-end encoder decoder configuration to develop detectors for two exemplary audio defects. A click detector is trained and compared to a traditional signal processing method, with a discussion on context sensitivity. Additional post-processing is used for data augmentation and workflow simulation. The ability of our models to capture variance is explored in a detector for artefacts from decompression of corrupted MP3 compressed audio. For both tasks we describe the synthetic generation of artefacts for controlled detector training and evaluation. We evaluate our detectors on the large open-source Free Music Archive (FMA) and genre-specific datasets.
Modeling polyphonic music is a particularly challenging task because of the intricate interplay between melody and harmony. A good model should satisfy three requirements: statistical accuracy (capturing faithfully the statistics of correlations at various ranges, horizontally and vertically), flexibility (coping with arbitrary user constraints), and generalization capacity (inventing new material, while staying in the style of the training corpus). Models proposed so far fail on at least one of these requirements. We propose a statistical model of polyphonic music, based on the maximum entropy principle. This model is able to learn and reproduce pairwise statistics between neighboring note events in a given corpus. The model is also able to invent new chords and to harmonize unknown melodies. We evaluate the invention capacity of the model by assessing the amount of cited, re-discovered, and invented chords on a corpus of Bach chorales. We discuss how the model enables the user to specify and enforce user-defined constraints, which makes it useful for style-based, interactive music generation.
Creating a good drum track to imitate a skilled performer in digital audio workstations (DAWs) can be a time-consuming process, especially for those unfamiliar with drums. In this work, we introduce PocketVAE, a groove generation system that applies grooves to users' rudimentary MIDI tracks, i.e, templates. Grooves can be either transferred from a reference track, generated randomly or with conditions, such as genres. Our system, consisting of different modules for each groove component, takes a two-step approach that is analogous to a music creation process. First, the note module updates the user template through addition and deletion of notes; Second, the velocity and microtiming modules add details to this generated note score. In order to model the drum notes, we apply a discrete latent representation method via Vector Quantized Variational Autoencoder (VQ-VAE), as drum notes have a discrete property, unlike velocity and microtiming values. We show that our two-step approach and the usage of a discrete encoding space improves the learning of the original data distribution. Additionally, we discuss the benefit of incorporating control elements - genre, velocity and microtiming patterns - into the model.
Diffusion probabilistic models (DPMs) have become a popular approach to conditional generation, due to their promising results and support for cross-modal synthesis. A key desideratum in conditional synthesis is to achieve high correspondence between the conditioning input and generated output. Most existing methods learn such relationships implicitly, by incorporating the prior into the variational lower bound. In this work, we take a different route -- we enhance input-output connections by maximizing their mutual information using contrastive learning. To this end, we introduce a Conditional Discrete Contrastive Diffusion (CDCD) loss and design two contrastive diffusion mechanisms to effectively incorporate it into the denoising process. We formulate CDCD by connecting it with the conventional variational objectives. We demonstrate the efficacy of our approach in evaluations with three diverse, multimodal conditional synthesis tasks: dance-to-music generation, text-to-image synthesis, and class-conditioned image synthesis. On each, we achieve state-of-the-art or higher synthesis quality and improve the input-output correspondence. Furthermore, the proposed approach improves the convergence of diffusion models, reducing the number of required diffusion steps by more than 35% on two benchmarks, significantly increasing the inference speed.