Finding temporal structure in music: blues improvisation with LSTM recurrent networks

By Douglas Eck, J. Urgen Schmidhuber. 2002

279 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Cognitive Neuroscience, Generative Models, Signal Processing

How researchers see this paper

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

In Artificial Intelligence

In Cognitive Neuroscience

In Generative Models

In Signal Processing

Citations in the corpus

Music Transformer: Generating Music with Long-Term Structure (2018, methodology): “Sequence models have been the canonical choice for modeling music, from Hidden Markov Models to RNNs and Long Short Term Memory networks (e.g., Eck & Schmidhuber, 2002; Liang, 2016; Oore et al., 2018), to bidirectional LSTMs (e.g., Hadjeres et al., 2017).”
Sequence Tutor: Conservative Fine-Tuning of Sequence Generation Models with KL-control (2016, background): “Often, this involves training the RNN to predict the next note in a monophonic melody; however, as mentioned above, the melodies generated by this model tend to wander and lack musical structure.”
Modeling Temporal Dependencies in High-Dimensional Sequences: Application to Polyphonic Music Generation and Transcription (2012, background): “While most existing models output only monophonic notes along with pre-deﬁned chords or other reduced-dimensionality representation (e.g. Mozer, 1994; Eck & Schmidhuber, 2002; Paiement et al., 2009), we aim to model unconstrained polyphonic music in the piano-roll representation, i.e. as a binary…”
Generating Focussed Molecule Libraries for Drug Discovery with Recurrent Neural Networks (2017, background): “[48] In this work, we highlight the analogy of language and chemistry, and show that RNNs can also generate reasonable molecules.”
Audio Chord Recognition with Recurrent Neural Networks (2013, background): “RNNs are completely general in that in principle they can describe arbitrarily complex long-term temporal dependencies, which made them very successful in music applications [5–7,11,23].”
Supervised Sequence Labelling with Recurrent Neural Networks (2012, background): “, 2007; Chen and Chaudhari, 2005), music generation (Eck and Schmidhuber, 2002), reinforcement learning (Bakker, 2002), speech recognition (Graves and Schmidhuber, 2005b; Graves et al.”
Unsupervised Incremental Online Learning and Prediction of Musical Audio Signals (2015, methodology): “Successfully applied to protein homology detection, automatic composition [48], handwriting and spoken language recognition, LSTM could be used to replace CB or HN and improve learning speed in our application.”
An End-to-End Neural Network for Polyphonic Piano Music Transcription (2015, background): “This makes them more suited for AMT tasks, since consecutive frames of audio exhibit both short-term and long-term temporal patterns [31].”
Novice-AI Music Co-Creation via AI-Steering Tools for Deep Generative Models (2020, background): “Research has demonstrated the potential for modeling and synthesizing music, ranging from single-voice sequences [13] and multi-part music [19, 34], to music with variable parts at each time step [4] and music with long-term structure over minutes [30, 37, 26].”
MidiNet: A Convolutional Generative Adversarial Network for Symbolic-Domain Music Generation (2017, background): “The majority of existing neural network models for music generation use recurrent neural networks (RNNs) and their variants, presumably for music generation is inherently about generating sequences [2,3,9,14].”
Recent Advances in Recurrent Neural Networks (2017, background): “LSTM models have since been adapted in music generation to better learn the long-term temporal structure of certain genres of music [116], [117].”
This time with feeling: learning expressive musical performance (2018, background): “ with papers on computational approaches to perception, interpretation and generation of music by Simon, Longuet-Higgins and others [23, 24, 25, 26, 36]. Since then, many interesting eﬀorts were made [15, 39, 30, 9, 34, 19], and it is clear that in recent years both interest and progress in score ge”
Understanding LSTM - a tutorial into Long Short-Term Memory Recurrent Neural Networks (2019, unknown): “Covered problems include protein secondary structure prediction [40, 9] and music generation [15].”
Travel time prediction with LSTM neural network (2016, background): “music generation [26], image captioning [27], speech recognition [28], machine translation [29].”
C-RNN-GAN: Continuous recurrent neural networks with adversarial training (2016, methodology): “Feeding the output from the previous cell is common practice when training RNNs as language models [Mikolov et al., 2010], and has also been used in music composition [Eck and Schmidhuber, 2002].”
Discovering Gated Recurrent Neural Network Architectures (2018, background): “); Eck and Schmidhuber ((2002)) demonstrated that these patterns are challenging to learn using traditional machine learning techniques.”