Eva Giboulot
YOU?
Author Swipe
View article: Evaluating the security of public surrogate watermark detectors
Evaluating the security of public surrogate watermark detectors Open
International audience
View article: WaterMax: breaking the LLM watermark detectability-robustness-quality trade-off
WaterMax: breaking the LLM watermark detectability-robustness-quality trade-off Open
Watermarking is a technical means to dissuade malfeasant usage of Large Language Models. This paper proposes a novel watermarking scheme, so-called WaterMax, that enjoys high detectability while sustaining the quality of the generated text…
View article: The Non-Zero-Sum Game of Steganography in Heterogeneous Environments
The Non-Zero-Sum Game of Steganography in Heterogeneous Environments Open
The highly heterogeneous nature of images found in real-world environments, such as online sharing platforms, has been one of the long-standing obstacles to the transition of steganalysis techniques outside the laboratory. Recent advances …
View article: Statistical Steganography based on a Sensor Noise Model using the Processing Pipeline
Statistical Steganography based on a Sensor Noise Model using the Processing Pipeline Open
Steganography is the discipline concerned with techniques designed to embed hidden data into an innocuous cover media. In the case of this manuscript, the cover media of choice are JPEG images. Steganography schemes based on a statistical …
View article: Multivariate Side-Informed Gaussian Embedding Minimizing Statistical Detectability
Multivariate Side-Informed Gaussian Embedding Minimizing Statistical Detectability Open
Steganography schemes based on a deflection criterion for embedding posses a clear advantage against schemes based on heuristics as they provide a direct link between theoretical detectability and empirical performance. However, this advan…
View article: Efficient Steganography in JPEG Images by Minimizing Performance of Optimal Detector
Efficient Steganography in JPEG Images by Minimizing Performance of Optimal Detector Open
International audience
View article: Detectability-Based JPEG Steganography Modeling the Processing Pipeline: The Noise-Content Trade-off
Detectability-Based JPEG Steganography Modeling the Processing Pipeline: The Noise-Content Trade-off Open
The current art of steganography shows that schemes using a deflection criterion (such as MiPOD) for JPEG steganography are usually subpar with respect to distortionbased schemes. We link this lack of performance to a poor estimation of th…
View article: Synchronization Minimizing Statistical Detectability for Side-Informed JPEG Steganography
Synchronization Minimizing Statistical Detectability for Side-Informed JPEG Steganography Open
Current schemes in steganography relying on synchronization are all based on a general heuristic to take into account interactions between embedding changes. However these approaches, while often competitive, lack a clear model for the rel…
View article: ALASKA#2: Challenging Academic Research on Steganalysis with Realistic Images
ALASKA#2: Challenging Academic Research on Steganalysis with Realistic Images Open
International audience
View article: Steganography by Minimizing Statistical Detectability: The cases of JPEG and Color Images
Steganography by Minimizing Statistical Detectability: The cases of JPEG and Color Images Open
International audience
View article: JPEG Steganography with Side Information from the Processing Pipeline
JPEG Steganography with Side Information from the Processing Pipeline Open
International audience
View article: Breaking ALASKA
Breaking ALASKA Open
This paper describes the architecture and training of detectors developed for the ALASKA steganalysis challenge. For each quality factor in the range 60-98, several multi-class tile detectors implemented as SRNets were trained on various c…
View article: The ALASKA Steganalysis Challenge
The ALASKA Steganalysis Challenge Open
International audience
View article: Steganalysis into the Wild: How to Define a Source?
Steganalysis into the Wild: How to Define a Source? Open
It is now well known that practical steganalysis using machine learning techniques can be strongly biased by the problem of Cover Source Mismatch. Such a phenomenon usually occurs in machine learning when the training and the testing sets …