Musical genre classification of audio signals

By George Tzanetakis, Perry R. Cook. IEEE Transactions on Speech and Audio Processing, 2002

965 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Signal Processing

View Paper: https://www.semanticscholar.org/paper/7ab881283270e427b05c6e9469562ff39dd6282a

How researchers see this paper

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

In Computer Vision and Pattern Recognition

In Signal Processing

Citations in the corpus

The Million Song Dataset (2011, methodology): “GZTAN [11]) while larger ones have resorted to solutions such as using songs released under Creative Commons (Magnatagatune [9]).”
Learning Features from Music Audio with Deep Belief Networks (2010, methodology): “The best reported accuracy on the Tzanetakis dataset [23] for genre classification was achieved by a system that used auditory cortical representations of music recordings and sparse representation-based classifiers [20].”
Aggregate features and ADABOOST for music classification (2006, methodology): “This includes evaluating features extracted from the beat histogram Tzanetakis and Cook (2002) and the autocorrelation phase matrix Eck and Casagrande (2005).”
A comparative study on content-based music genre classification (2003, methodology): “To the best of our knowledge, currently the most influential approach to direct modeling of music signals for automatic genre classification is due to Tsanetakis and Cook [31], where the timbral texture, rhythm, and pitch content features are explicitly developed.”
Gender classification in two Emotional Speech databases (2008, methodology): “A system using GMMs is proposed by Tzanetakis and Cook that achieves 74% accuracy [11].”
Multi-Pitch Estimation (2009, background): “, [170]), and the parametric representations used in this book may, therefore, directly or indirectly, be used for this.”
Qwen3-Omni Technical Report (2025, positive): “On GTZAN, MTG-Jamendo, and MagnaTagATune, the scores of Qwen3-Omni-Instruct also significantly surpass other audio language models, including Gemini-2.”
Factors in automatic musical genre classification of audio signals (2003, methodology): “Means and variances of the features over a larger texture window (1 second) with a hop size of 20 milliseconds are computed resulting in a set of 18 features.”
Developing crossmodal expression recognition based on a deep neural model (2016, background): “The GTZAN corpus by Tzanetakis and Cook (2002) is not directly related to emotion recognition, but to music genre classification.”
Learning-Based Methods for Comparing Sequences, with Applications to Audio-to-MIDI Alignment and Matching (2016, methodology): “For our baseline, we followed the tradition of (Tzanetakis and Cook, 2002; Deng et al., 2008; Mandel and Ellis, 2005; Foster et al., 2014) and others by summarizing spectrograms by their mean and standard deviation across feature dimensions and concatenating the resulting statistics into a fixed-len”
Multi-Label Classification of Music into Emotions (2008, methodology): “In terms of features, they used Marsyas [16] to extract 30 features related to the timbral texture, rhythm and pitch.”