Binary Non-Negative Matrix Deconvolution for Audio Dictionary Learning
| Autor: | Szymon Drgas, Jörg Lücke, Antti Hurmalainen, Tuomas Virtanen |
|---|---|
| Přispěvatelé: | Tampere University, Signal Processing, Research group: Audio research group - ARG |
| Jazyk: | angličtina |
| Rok vydání: | 2017 |
| Předmět: |
Audio signal
Acoustics and Ultrasonics Computer science business.industry Speech recognition Vector quantization Binary number 020206 networking & telecommunications Pattern recognition 02 engineering and technology 113 Computer and information sciences 030507 speech-language pathology & audiology 03 medical and health sciences Computational Mathematics Matrix (mathematics) Encoding (memory) 0202 electrical engineering electronic engineering information engineering Computer Science (miscellaneous) Spectrogram Deconvolution Nonnegative matrix Artificial intelligence Electrical and Electronic Engineering 0305 other medical science business |
| Popis: | In this study, we propose an unsupervised method for dictionary learning in audio signals. The new method, called binary nonnegative matrix deconvolution (BNMD), is developed and used to discover patterns from magnitude-scale spectrograms. The BNMD models an audio spectrogram as a sum of delayed patterns having binary gains (activations). Only small subsets of patterns can be active for a given spectrogram excerpt. The proposed method was applied to speaker identification and separation tasks. The experimental results show that dictionaries obtained by the BNMD bring much higher speaker identification accuracies averaged over a range of SNRs from -6 dB to 9 dB (91.3%) than the NMD-based dictionaries (37.8-75.4%). The BNMD also gives a benefit over dictionaries obtained using vector quantization (87.8%). For bigger dictionaries the difference between the BNMD and the vector quantization (VQ) is getting smaller. For the speech separation task the BNMD dictionary gave a slight improvement over the VQ. acceptedVersion |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|