Frustratingly Easy Edit-based Linguistic Steganography with a Masked Language Model
TL
TL;DR
The proposed method eliminates painstaking rule construction and has a high payload capacity for an edit-based model, and is shown to be more secure against automatic detection than a generation-based method while offering better control of the security/payload capacity trade-off.
Abstract
With advances in neural language models, the focus of linguistic steganography has shifted from edit-based approaches to generation-based ones. While the latter’s payload capacity is impressive, generating genuine-looking texts remains challenging. In this paper, we revisit edit-based linguistic steganography, with the idea that a masked language model offers an off-the-shelf solution. The proposed method eliminates painstaking rule construction and has a high payload capacity for an edit-based model. It is also shown to be more secure against automatic detection than a generation-based method while offering better control of the security/payload capacity trade-off.
References26 items
1
Near-imperceptible Neural Linguistic Steganography via Self-Adjusting Arithmetic Coding
2
Adversarial Watermarking Transformer: Towards Tracing Text Provenance with Data Hiding
3
Linformer: Self-Attention with Linear Complexity
4
Longformer: The Long-Document Transformer
5