Hide and Speak: Towards Deep Neural Networks for Speech Steganography
This work addresses the need for secure and imperceptible speech steganography, offering a novel deep learning-based approach that improves upon existing methods for this specific domain.
The paper tackled the problem of hiding secret messages in speech data using deep neural networks, proposing a model that incorporates short-time Fourier transforms as differentiable layers to impose constraints, and demonstrated its effectiveness with quantitative and qualitative results on several speech datasets.
Steganography is the science of hiding a secret message within an ordinary public message, which is referred to as Carrier. Traditionally, digital signal processing techniques, such as least significant bit encoding, were used for hiding messages. In this paper, we explore the use of deep neural networks as steganographic functions for speech data. We showed that steganography models proposed for vision are less suitable for speech, and propose a new model that includes the short-time Fourier transform and inverse-short-time Fourier transform as differentiable layers within the network, thus imposing a vital constraint on the network outputs. We empirically demonstrated the effectiveness of the proposed method comparing to deep learning based on several speech datasets and analyzed the results quantitatively and qualitatively. Moreover, we showed that the proposed approach could be applied to conceal multiple messages in a single carrier using multiple decoders or a single conditional decoder. Lastly, we evaluated our model under different channel distortions. Qualitative experiments suggest that modifications to the carrier are unnoticeable by human listeners and that the decoded messages are highly intelligible.