A Comparative Re-Assessment of Feature Extractors for Deep Speaker Embeddings
Tiedosto(t)
Rinnakkaistallenteen versio
published versionPäivämäärä
2020Tekijä(t)
Yksilöllinen tunniste
10.21437/Interspeech.2020-1765Metadata
Näytä kaikki kuvailutiedotLisätietoa
Rinnakkaistallenne
Viittaus
Liu, Xuechen. Sahidullah, Md. Kinnunen, Tomi. (2020). A Comparative Re-Assessment of Feature Extractors for Deep Speaker Embeddings. Proceedings of Interspeech 2020, 2020, 3221-3225. 10.21437/Interspeech.2020-1765.Oikeudet
Tiivistelmä
Modern automatic speaker verification relies largely on deep neural networks (DNNs) trained on mel-frequency cepstral coefficient (MFCC) features. While there are alternative feature extraction methods based on phase, prosody and long-term temporal operations, they have not been extensively studied with DNN-based methods. We aim to fill this gap by providing extensive re-assessment of 14 feature extractors on VoxCeleb and SITW datasets. Our findings reveal that features equipped with techniques such as spectral centroids, group delay function, and integrated noise suppression provide promising alternatives to MFCCs for deep speaker embeddings extraction. Experimental results demonstrate up to 16.3% (VoxCeleb) and 25.1% (SITW)
relative decrease in equal error rate (EER) to the baseline.