CaSPR Lab – Human Distinguishable Visual Key Fingerprints

Image of Human Perception Discriminator (HPD) architecture

The CaSPR labs has released CEAL, a visual key fingerprint generation system that enables users to easily compare arbitrarily complex and long strings (e.g. cryptographic keys, addresses, identifiers) received from a trusted and an untrusted source, by converting it to the equivalent yet more natural task of comparing images or shorter text (i.e. fingerprints). CEAL has applications to manual key verification in SSH, OpenPGP and end-to-end encrypted (E2EE) applications such as WhatsApp, Viber, LinkedIn, the detection and prevention of phishing and Bitcoin clipboard attacks that target millions of addresses, device pairing, and the development of visual CAPTCHAs. This work was accepted in the Usenix Security Symposium and will be presented in August 2020. 

Visit here to read the paper: http://www.cs.fiu.edu/~carbunar/ceal.pdf