Cryptography meets computer vision.
Pixek is a camera app that protects you against data breaches and unwanted photo disclosures.
Structured Encryption is an advanced form of encryption designed to support searching on encrypted data.
When you search over your encrypted photo collection, the cloud only sees an encrypted query but is still able to return the relevant encrypted photos. The only information we learn when you search for something is whether you searched for the same thing in the past; but not what you searched for.
Pixek uses convolutional neural nets to automatically tag your pictures with their content. The tags and photos are encrypted on your phone before being sent to the cloud.
Photos have high sentimental value, and many of us have personal reasons for wanting to keep our photos private.
We may be uncomfortable with companies having access to pictures of our families, or we may want to backup photos with confidential information like work-related documents. We may be afraid of data breaches and photo leaks.
A picture is worth a thousand words and photography is an important part of any news story. The preservation of confidentiality is important to protecting sources, avoiding early leaks of stories, or simply evading censure.
Smartphones have become an important tool for activism throughout the world. User-generated pictures have exposed government abuses and in war-areas citizen-journalists are one of the few reliable and independent sources of information left. These photos must be protected so that the world can see them.
Martin Zhu is a software engineer. He holds a degree in Computer Science from Brown University and has extensive experience building consumer applications from the ground up. On top of his engineering background, he has worked on various cryptography, deep learning, and Big Data research projects.
Tarik Moataz is a postdoctoral research fellow at Brown University. He holds a PhD from Institut Mines-Telecom Atlantique and Colorado State University. His research includes several advances in encrypted algorithms and systems. He built and maintains the Clusion library, a state-of-the-art open source encrypted search library.
Seny Kamara is a professor of Computer Science at Brown University. He holds a PhD from Johns Hopkins University and has worked on encrypted search for over a decade. His research has contributed to the foundations and state-of-the-art of encrypted search algorithms and their cryptanalysis.