Face Recognition for Authenticating Network Transactions

As software agents begin to help people automatically manage information access and distribution across networks, authenticating transactions will become increasingly important. To the extent that agent interactions will have similar properties to people interactions, we can begin to predict some of the issues that will become important. One key issue is trust in transactions. People need to trust that expectations made for a transaction will come true. At the most fundamental level, is the trust that someone is who they claim to be or that an agent really represents the right person.

Today we use passwords and cards to validate users. Unfortunately, they can be lost, stolen or forged. The use of biometrics like fingerprints, voice and face increases the level of trust for authentication. And the more trust people have in safe transactions, the more they will use and rely on them. We believe that face recognition will play an especially key role in user authentication for a number of reasons: Face recognition is more reliable than passwords, less-intrusive than hand geometry, more convenient than voice recognition, rejects fewer authorized people than fingerprints, fast and reasonably priced.

At the forum, I will discuss various applications for face recognition in both verifying and identifying users as well as the role that face recognition might play for creating trust in network transactions.

Michael Kuperstein

CEO, Miros, Inc.

Back to FIPA Opening Forum program