Adversarial Representation Learning for Synthetic Replacement of Sensitive Speech Data

Typ
Examensarbete för masterexamen
Program
Engineering mathematics and computational science (MPENM), MSc
Publicerad
2020
Författare
Östberg, Adam
Ericsson, David
Modellbyggare
Tidskriftstitel
ISSN
Volymtitel
Utgivare
Sammanfattning
As more data is collected in various settings across organizations, companies, and countries, there has been an increase in the demand of user privacy. Developing privacy preserving methods for data analytics is thus an important area of research. In this work we present a model based on generative adversarial networks (GANs) that learns to obfuscate specific sensitive attributes in speech data. We train a model that learns to hide sensitive information in the data, while preserving the meaning in the utterance. The model is trained in two steps: first to filter sensitive information in the spectrogram domain, and then to generate new and private information independent of the filtered one. The model is based on a CNN that takes mel-spectrograms as input. A MelGAN is used to invert the spectrograms back to raw audio waveforms. We show that it is possible to hide sensitive information such as gender by generating new data, trained adversarially to maintain utility and realism.
Beskrivning
Ämne/nyckelord
generative adversarial networks, adversarial representation learning, deep learning, privacy, speech generation
Citation
Arkitekt (konstruktör)
Geografisk plats
Byggnad (typ)
Byggår
Modelltyp
Skala
Teknik / material
Index