A First Look at Music Composition using LSTM Recurrent Neural Networks

By Douglas Eck, J. Urgen Schmidhuber. 2002

287 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Generative Models

Citations in the corpus

Generating Sequences With Recurrent Neural Networks (2013, background): “Recurrent neural networks (RNNs) are a rich class of dynamic models that have been used to generate sequences in domains as diverse as music [6, 4], text [30] and motion capture data [29].”
Molecular de-novo design through deep reinforcement learning (2017, methodology): “The Recurrent Neural Network (RNN) is commonly used as a generative model for data of sequential nature, and have been used successfully for tasks such as natural language processing [11] and music generation [12].”
Memory-Efficient Backpropagation Through Time (2016, background): “LSTM [9] in particular is an RNN architecture that has excelled in sequence generation [3, 13, 4], speech recognition [5] and reinforcement learning [12, 10] settings.”
Deep Bidirectional and Unidirectional LSTM Recurrent Neural Network for Network-wide Traffic Speed Prediction (2018, methodology): “…Traffic Speed Prediction Zhiyong Cui, Ruimin Ke, Ziyuan Pu, Yinhai Wang T data, LSTMs have been proved to be able to process sequence data [4] and applied in many real-world problems, like speech recognition [6], image captioning [7], music composition [8] and human trajectory prediction [9].”
Predict Stock Price with Financial News Based on Recurrent Convolutional Neural Networks (2017, background): “The LSTM architecture [11] [12][13] is composed of a collection of recurrently connected chained networks, usually known as memory blocks.”
Deep Learning Techniques for Music Generation - A Survey (2017, methodology): “In Eck and Schmidhuber’s second experiment [37], the objective is to simultaneously generate melody and chord sequences.”
A Functional Taxonomy of Music Generation Systems (2017, background): “…and Schmidhuber 2002; Boulanger-Lewandowski et al. 2012; Herremans and Chuan 2017] Melody [Todd 1989; Duff 1989; Mozer 1991; Lewis 1991; Toiviainen 1995; Eck and Schmidhuber 2002; Franklin 2006; Agres et al. 2009 Melody constitutes one of the ﬁrst aspects of music subject to automatic generation.”
Sequence to Better Sequence: Continuous Revision of Combinatorial Structures (2017, background): “The success of recurrent neural network (RNN) models in complex tasks like machine translation and audio synthesis has inspired immense interest in learning from sequence data (Eck & Schmidhuber, 2002; Graves, 2013; Sutskever et al., 2014; Karpathy, 2015).”
A Survey on Modern Deep Neural Network for Traffic Prediction: Trends, Methods and Challenges (2020, methodology): “Due to this, RNN based models have been applied in many fields such as named entity recognition [40], voice recognition [41], music composition [42], and image caption generation [43].”
Evolving Recurrent Neural Networks for Time Series Data Prediction of Coal Plant Parameters (2019, methodology): “LSTM RNNs have been used with strong performance in image recognition [8], audio visual emotion recognition [9], music composition [10] and other areas.”
A Hierarchical Recurrent Neural Network for Symbolic Melody Generation (2017, methodology): “As an important form of RNN, LSTM (Hochreiter and Schmidhuber 1997) is used to capture the global music structure and improve the quality of the generated music (Eck and Schmidhuber 2002).”
Imposing higher-level Structure in Polyphonic Music Generation using Convolutional Restricted Boltzmann Machines and Constraints (2016, methodology): “More recently Eck and Schmidhuber (2002) trained a Long Short-term Memory (LSTM) network (a state-of-the-art RNN variant), jointly on a single chord sequence along with several different melodies. is an example of a harmonic template which guides a melodic improvisation.”
MUlti-Store Tracker (MUSTer): A cognitive psychology inspired approach to object tracking (2015, unknown): “speech recognition [13] and music composition [10].”
Song From PI: A Musically Plausible Network for Pop Music Generation (2016, methodology): “It brings together machines learning researchers that aim to capture the complex structure of music (Eck & Schmidhuber (2002); Boulanger-lewandowski et al. (2012)), as well as music professionals (Chan et al. (2006)) and en-thusiasts (Johnson; Sun) that want to see how far a computer can get to be…”
Recent Advances in Recurrent Neural Networks (2017, background): “It has been shown that RNN models struggle to keep track of distant events that indicate the temporal structure of music [116].”