DeepSpectrum is a Python toolkit for feature extraction from audio data with pre-trained Image Convolutional Neural Networks (CNNs). It features an extraction pipeline which first creates visual representations for audio data - plots of spectrograms or chromagrams - and then feeds them to a pre-trained Image CNN. Activations of a specific layer then form the final feature vectors.

(c) 2017-2020 Shahin Amiriparian, Maurice Gerczuk, Sandra Ottl, Björn Schuller: Universität Augsburg Published under GPLv3, see the LICENSE.md file for details.

Please direct any questions or requests to Shahin Amiriparian (shahin.amiriparian at tum.de) or Maurice Gercuk (maurice.gerczuk at informatik.uni-augsburg.de).

Citing

If you use DeepSpectrum or any code from DeepSpectrum in your research work, you are kindly asked to acknowledge the use of DeepSpectrum in your publications.

S. Amiriparian, M. Gerczuk, S. Ottl, N. Cummins, M. Freitag, S. Pugachevskiy, A. Baird and B. Schuller. Snore Sound Classification using Image-Based Deep Spectrum Features. In Proceedings of INTERSPEECH (Vol. 17, pp. 2017-434)