Subjective Intelligibility of Deep Neural Network-Based Speech Enhancement
| Autor: | Erlend Magnus Viggen, Tron Vedul Tronstad, Femke B. Gelderblom |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2017 |
| Předmět: |
Artificial neural network
Computer science business.industry Speech recognition 020206 networking & telecommunications 02 engineering and technology Intelligibility (communication) computer.software_genre Speech enhancement 030507 speech-language pathology & audiology 03 medical and health sciences 0202 electrical engineering electronic engineering information engineering Artificial intelligence 0305 other medical science business computer Natural language processing |
| Zdroj: | Interspeech INTERSPEECH |
| Popis: | Recent literature indicates increasing interest in deep neural networks for use in speech enhancement systems. Currently, these systems are mostly evaluated through objective measures of speech quality and/or intelligibility. Subjective intelligibility evaluations of these systems have so far not been reported. In this paper we report the results of a speech recognition test with 15 participants, where the participants were asked to pick out words in background noise before and after enhancement using a common deep neural network approach. We found that, although the objective measure STOI predicts that intelligibility should improve or at the very least stay the same, the speech recognition threshold, which is a measure of intelligibility, deteriorated by 4 dB. These results indicate that STOI is not a good predictor for the subjective intelligibility of deep neural network-based speech enhancement systems. We also found that the postprocessing technique of global variance normalisation does not significantly affect subjective intelligibility. Copyright © 2017 ISCA |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|