Jessica Fridrich
Binghamton University, NY·Computer Science
Steganography hides a secret message inside an innocent-looking object, such as a digital photo, by changing the image so slightly that neither people nor computers can tell. It can offer privacy where encryption is banned.
- Fields
Computer Science
Engineering
Mathematics & Statistics
- Topics
steganography
steganalysis
information hiding
- Grants
- 1 known grant, through Sep 2027
What they work on
- Use artificial but realistic image datasets to compute theoretical limits of steganography and assess how close modern data-driven hiding algorithms and steganalysis detectors come to optimal.
- Train machine learning models restricted in task, training-example format and architecture to discover how AI detectors spot hidden messages in digital images.
- Build new analytic models of steganography and its detection from the insights of restricted machine learning models, giving practitioners explainable tools.
More about the research
Criminals, terrorists and spies can also misuse it. Today both hiding messages and detecting them, a task called steganalysis, rely heavily on machine learning. These "black box" systems outperform older analytical methods, but little is known about their performance limits or why they work.
Why it matters
Understanding the limits of hidden communication in digital images could support better defenses against steganography used by terrorists, criminals and spies to coordinate disruptive activities. The resulting methods may also be useful in related fields such as digital forensics and data provenance, which trace where digital content came from.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Jie Yan
Bowie State University, MD·Computer Science
Investigate the sources of hallucination — fabricated or inaccurate output — when large language models are applied to digital forensic casework, and develop techniques that reduce it enough for results to align with evidentiary standards.
Matthew Wright
Rochester Institute of Technology, NY·Computer Science
Study how users evaluate suspect digital content through qualitative interviews and user studies, and use the findings to integrate multiple AI-content forensic tools into one navigable platform.
Siwei Lyu
University at Buffalo, NY·Computer Science
Strengthen deepfake detectors against anti-forensic attacks by processing each input with a randomly chosen model from an infinite ensemble of classifiers, making detectors harder to fool.
The school
Binghamton University: 14 labs on PI Pages, $161M a year on research, 132 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.