Apr 6, 2023 · In this paper, we propose Disentangled Spectrogram Variational Auto Encoder (DSVAE) which is a two staged trained variational autoencoder that processes ...

Disentangled Spectrogram Variational Auto Encoder (DSVAE) is a two staged trained variational autoencoder that processes spectrograms of speech using ...

Our experimental results show high accuracy (> 98%) on detecting synthetic speech from 6 known and 10 unknown speech synthesizers. Further, the visualization of ...

Co-authors ; Dsvae: Interpretable disentangled representation for synthetic speech detection. AKS Yadav, K Bhagtani, Z Xiang, P Bestagini, S Tubaro, EJ Delp.

In this paper, we propose Disentangled Spectrogram Variational Auto Encoder (DSVAE) which is a two staged trained variational autoencoder that processes ...

... Disentangled representation learning methods leverage the idea that it is possible to divide learned representations into interpretable components [12, 70].

... Disentangled representation learning methods leverage the idea that it is possible to divide learned representations into multiple explainable components [ ...

In this work, we examine bias in existing synthetic speech detectors to determine if they will unfairly target a particular gender, age and accent group.

DSVAE: Disentangled Representation Learning for Synthetic Speech Detection. AKS Yadav, K Bhagtani, Z Xiang, P Bestagini, S Tubaro, EJ Delp. 2023 International ...

DSVAE also creates an activation map to highlight the spectrogram regions that discriminate synthetic and bona fide human speech signals. We evaluated the ...