Deep Learning Papers on Music Analysis

[u][b]Papers on Deep Learning & Music IR[/b][/u]

• Moving beyond feature design: Deep architectures and automatic feature learning in music informatics. E. Humphrey, J. Bello, and Y. LeCun. In: Proceedings of International Symposium on Music Information Retrieval (ISMIR), 2012

• Feature learning and deep architectures: New directions for music informatics. E. Humphrey, J. Bello, and Y. LeCun. Journal of Intelligent Information Systems, 2013

• Latin genre recognition with deep learning (https://highnoongmt.wordpress.com/2015/06/22/el-caballo-viejo/)

• El Caballo Viejo Latin Genre Recognition with Deep Learning and Spectral Periodicity(http://scholar.google.com/scholar_url?url=http://www2.imm.dtu.dk/pubdb/views/edoc_download.php/6880/pdf/imm6880.pdf&hl=en&sa=X&scisig=AAGBfm3YQecZ_L3g6BeEiipUcH95sA1zbg&nossl=1&oi=scholaralrt) BL Sturm, C Kireliuk, J Larsen - Fifth Biennial International Conference on Mathematics … Abstract. The “winning” system in the 2013 MIREX Latin Genre Classification Train-test Task was a deep neural network trained with simple periodicity features. The explanation for its winning performance has yet to be fully explained. In our previous work, we built similar ...

• Deep Learning and Music Adversaries(http://scholar.google.at/scholar_url?url=http://arxiv.org/pdf/1507.04761&hl=de&sa=X&scisig=AAGBfm2hgm8ATPVboTMJMlx3Z9Q41svx7g&nossl=1&oi=scholaralrt) C Kereliuk, BL Sturm, J Larsen - arXiv preprint arXiv:1507.04761, 2015 Abstract: An adversary is essentially an algorithm intent on making a classification system perform in some particular way given an input, eg, increase the probability of a false negative. Recent work builds adversaries for deep learning systems applied to image …

• A Deep Neural Network for Modeling Music(http://scholar.google.com/scholar_url?url=http://homepage.fudan.edu.cn/zhengxq/files/2015/04/ICMR2015.pdf&hl=de&sa=X&scisig=AAGBfm0v0GeRf7lnqqS55D45wI2DuPVWuQ&nossl=1&oi=scholaralrt) P Zhang, X Zheng, W Zhang, S Li, S Qian, W He… - 2015 ... systems. Extensive research effort (see 5, 3, 27, 15 for reviews) has been invested in content- based music information retrieval (MIR) at the intersection of signal processing, music modeling, and machine learn- ing. Although ...

• Unsupervised feature learning for audio classification using convolutional deep belief networks (http://papers.nips.cc/paper/3674-unsupervised-feature-learning-for-audio-classification-using-convolutional-deep-belief-networks.pdf)

• Unsupervised Feature Learning for Urban Sound Classification Justin Salamon http://www.justinsalamon.com/news/unsupervised-feature-learning-for-urban-sound-classification

• Generating Music Playlists with Hierarchical Clustering and Q-Learning (simple approach with skipping behaviour feedback) James King and Vaiva Imbrasaite (ECIR 2015) Www.james.eu.org/musicplayer [email protected]

• Deep Network Geometry of Timbre - A Geometric Approach to Timbre using deep neural networks Workshop on Musical Timbre, Télécom ParisTech – November 14, 2014 http://musictimbre.wp.mines-telecom.fr/files/2014/11/Mallat2014.pdf http://musictimbre.wp.mines-telecom.fr/keynotes/

• Semantic hashing. Salakhutdinov, R. and Hinton, G. (2009). International Journal of Approximate Reasoning. A crucial aspect of the learned hash codes is that semantically related items have similar hash codes (in terms of their Hamming distance). lrn2 provides a method to create a database from given data sets, such that data instances are indexed by their learned representation (in binarized form). http://www.sciencedirect.com/science/journal/0888613X/50/7

• Learning Binary Codes for Efficient Large-Scale Music Similarity Search Jan Schlüter, OFAI Presentation: http://www.ofai.at/~maarten.grachten/downloads/lrn2cre8/dlmp-workshop/slides/jan/autoencoder_binarycodes.pdf

• Isolated instrument transcription using a deep belief network(http://scholar.google.com/scholar_url?url=https://peerj.com/preprints/1193.pdf&hl=de&sa=X&scisig=AAGBfm3mrtrQDS3C6KbiTvZ-ZdnnVs9Jaw&nossl=1&oi=scholaralrt) G Burlet, A Hindle - 2015 ... However, these applications seem far out of grasp given that the music information retrieval (MIR) research community has collectively reached a plateau in the accuracy of automatic music transcription systems 3. In a paper addressing this issue, Benetos et al. ...

• Deep Remix: Remixing Musical Mixtures Using a Convolutional Deep Neural Network(http://scholar.google.com/scholar_url?url=http://arxiv.org/pdf/1505.00289&hl=de&sa=X&scisig=AAGBfm1a3mAOKRPrKMepXseqDeDKegYilw&nossl=1&oi=scholaralrt) AJR Simpson, G Roma, MD Plumbley - arXiv preprint arXiv:1505.00289, 2015 ... multitrack dataset for annotation-intensive MIR research”, In 15th Int. Soc. Music Information Retrieval Conf. 13 Simpson AJR (2015) “Abstract Learning via Demodulation in a Deep Neural Network”, arxiv.org abs/1502.04042.

• Deep Karaoke: Extracting Vocals from Musical Mixtures Using a Convolutional Deep Neural Network(http://scholar.google.com/scholar_url?url=http://arxiv.org/pdf/1504.04658&hl=de&sa=X&scisig=AAGBfm1GE8JnnL02ZeyWvhxPr7Sdh89rcw&nossl=1&oi=scholaralrt) AJR Simpson, G Roma, MD Plumbley - arXiv preprint arXiv:1504.04658, 2015 ... research”, In 15th Int. Soc. Music Information Retrieval Conf. 11 Terrell MJ, Simpson AJR, Sandler M (2014) “The Mathematics of Mixing”, Journal of the Audio Engineering Society, 62(1/2), 4-13. 12 Simpson AJR (2015) “Abstract ...

• Be at Odds? Deep and Hierarchical Neural Networks for Classification and Regression of Conflict in Speech(http://scholar.google.com/scholar_url?url=http://link.springer.com/chapter/10.1007/978-3-319-14081-0_19&hl=en&sa=X&scisig=AAGBfm07wZf8jXgWlwbW8W4TUxVXngyXtw&nossl=1&oi=scholaralrt) R Brueckner, B Schuller - Conflict and Multimodal Communication, 2015 Abstract Conflict is a fundamental phenomenon inevitably arising in inter-human communication and only recently has become the subject of study in the emerging field of computational paralinguistics. As speech is a predominant carrier of information about the ...

More papers on Deep Learning for Sound and Music Processing Workshop by OFAI http://lrn2cre8.eu/?q=deeplearningworkshop

[u][b]Deep Learning from Wave Forms[/b][/u]

• modelling speech waveforms with neural nets (Heiga Zen) (http://static.googleusercontent.com/media/research.google.com/es//pubs/archive/43267.pdf)

*Sander Dieleman: learning directly from waveforms. Comparison with spectrograms (https://dl.dropboxusercontent.com/u/19706734/paper_pt.pdf)

• PhD Thesis: Modeling High-Dimensional Audio Sequences with Recurrent Neural Networks(http://www-etud.iro.umontreal.ca/~boulanni/NicolasBoulangerLewandowski_thesis.pdf), Nicolas Boulanger-Lewandowski