About the Project

Built on Science.
Trusted by Evidence.

eWitness is powered by SmartHash — peer-reviewed research from John Jay College of Criminal Justice, developed in collaboration with The Graduate Center, CUNY and NYU.

The Problem We Solve

In an era of deepfakes, generative AI imagery, and rampant photo manipulation, the simple question "is this real?" has never been harder to answer. Misinformation and disinformation are escalating threats to journalism, public safety, and trust in democratic institutions.

Verifying digital images from unknown sources has become an essential part of the editorial process in the news industry. Social media platforms bear a corporate responsibility to establish trust in authentic content, recognize tampering early, and prevent circulation of images that do not come from verified sources.

eWitness addresses this challenge by registering a SmartHash fingerprint of every photo or video on an immutable blockchain at the moment of capture — making the chain of custody permanent and publicly verifiable.

"Content provenance can be built into a camera by leveraging a public distributed ledger to create an immutable record of the metadata and a hash of the image at the time of capture."
— Samanta & Jain, ACM CIKM 2024
Research Team

The People Behind SmartHash

P

Priyanka Samanta

Lead Researcher

The Graduate Center, CUNY & Brooklyn College

S

Shweta Jain

Principal Investigator

John Jay College of Criminal Justice & The Graduate Center, CUNY

The team acknowledges Prof. Nasir Memon, Professor and Dean of Engineering at NYU Shanghai, for his guidance and advice on the SmartHash research. This work was supported by the NSF TIP Directorate, Grant Number 2122682.

Institutional Partners

Where the Research Lives

⚖️

John Jay College of Criminal Justice

Lead Institution

Home of the principal research team. John Jay is a CUNY institution dedicated to the study of criminal justice, forensics, and public safety.

Visit Website →
🎓

The Graduate Center, CUNY

Research Partner

Doctoral students and faculty from the Graduate Center contributed to the theoretical foundations and experimental evaluation of SmartHash.

Visit Website →
🏛️

New York University

Collaborating Institution

Faculty from NYU provided expert guidance and advice on the algorithm design and evaluation methodology.

Visit Website →
Peer-Reviewed Publication

Published at ACM CIKM 2024

ACM

SmartHash: Perceptual Hashing for Image Tampering Detection and Authentication

Priyanka Samanta & Shweta Jain

CIKM '24: Proceedings of the 33rd ACM International Conference on Information and Knowledge Management · October 21–25, 2024 · Boise, ID, USA

DOI: 10.1145/3627673.3679827

3 Citations 471 Downloads Open Access

SmartHash is evaluated on five large tampered-image datasets and shown to have high accuracy and precision in detecting content-changing modifications. It outperforms Apple's NeuralHash and Microsoft's PhotoDNA for image authentication tasks while running 6–56× faster on single-core hardware — making it uniquely suitable for deployment on smartphones and resource-constrained devices.

Read the Full Paper →
Our Principles

What We Stand For

🔓

Open Research

SmartHash is published peer-reviewed open access. Our methods, datasets, and evaluation code are available for inspection, replication, and further research.

🛡️

Privacy by Design

eWitness never stores your media, your identity, or your location. Only the SmartHash fingerprint is registered — never the image itself.

🌍

Accessible to All

eWitness is free to download and use. Forensic-grade media authentication should not be gated by income, resources, or technical expertise.

🎓

Academic Rigor

Every claim in eWitness is backed by peer-reviewed research. SmartHash was evaluated on 5 public datasets totaling over 180,000 images before deployment.

Use the Science

Download eWitness and put peer-reviewed SmartHash authentication in your pocket — free.