I'm a research scientist working at the intersection of computer vision, signal processing and security. A central theme of my work to date focused on verification and assurance of media integrity - both in terms of pro-active secure media capture and passive media forensics. This interdisciplinary field combines concepts from computer vision, image processing, image quality assessment, machine learning, statistical signal processing, and communication & information theory.

If you’d like to know more about the field, feel free to check out my recent survey. If you'd like to experiment with some of the techniques that I was working on, the best place to start is the featured content below and my github profile. The publications section also lists supplementary materials and source code that accompanies specific papers.

I'm currently an applied scientist at Amazon. Previously, I worked with prof. Nasir Memon at NYU Center for Cyber-security on modeling and end-to-end optimization of photo acquisition and distribution pipelines. I did my post-doctoral research on photo manipulation detection and forensic analysis of imaging sensors with prof. Jiwu Huang at the Key Laboratory of Media Information Content Security at Shenzhen University, China. My doctoral work was under the supervision of prof. Andrzej Dziech at the Department of Telecommunications of AGH University of Science and Technology, Kraków, Poland. I used tools from communication and information theory to design new techniques for photo authentication and restoration based on digital watermarking.

I currently serve the research community as a reviewer for technical journals and conferences and as a member of the IEEE Information Forensics and Security Technical Committee.

Research Interests

low-level vision · machine learning · image processing · signal processing · media forensics · media integrity

Featured Content

computational sensor fingerprints

Learning next-generation camera fingerprints using neural sensors

tifs'22

dictionary attacks on speaker verification

Adversarial attacks maximizing impersonation capabilities of speech

interspeech'19 · arxiv'22

neural imaging toolbox

Modeling and optimization of complex photo acquisition and distribution channels

cvpr'19 · arxiv'19 · iclr'20 · tifs'22

lightweight learned lossy image codec

Lightweight codec with good rate-distortion performance (~BPG) and fast runtime

iclr'20

multi-scale tampering localization

Improved tampering localization thanks to multi-scale analysis and decision fusion

tip'16 · wifs'16 · tifs'17 · arxiv'18

photo integrity survey

Overview of techniques for verification and active protection of digital photographs

dsp'17

realistic tampering dataset

220 realistic forgeries created by hand in modern photo-editing software

wifs'16 · tifs'17

self-recovery of digital photos

Detailed study and practical demonstration of various edit recovery schemes

tip'13 · tifs'14 · wifs'14 · tmm'15

Media Coverage

• wired.com: To Fight Deepfakes, Researchers Built a Smarter Camera
• gizmodo.com: Researchers Want to Build Fake Photo Detection Tools Right Into Our Cameras
• photonics.com: AI-Driven Imaging System Protects Photo and Video Authenticity End-to-End
• verdict.co.uk: Fighting fake news with anti-deepfake technology
• helpnetsecurity.com: Researchers spot manipulated photos and video using AI-driven imaging system
• engineering.nyu.edu: Outsmarting deep fakes: Researchers devise an AI-driven imaging system that protects authenticity
• pdnonline.com: Your Future Camera May Watermark Images Automatically, Whether You Want It To or Not