For the past two decades, researchers have been working to fulfil the promise of ECG biometrics. While face can be stolen at a distance and fingerprints are left everywhere, the electrocardiogram would be the hidden secret to fight spoofing and counterfeiting. We just needed to develop systems as accurate, robust, and unobtrusive as those available for more developed traits. However, after twenty years and plenty of literature, we still aren’t there yet. Acquisition setups have never been more comfortable, deep learning brought us the most sophisticated models, and databases are unprecedentedly large and complete. So, what’s missing? This is the question João aims to answer in this talk, as he walks through the history of ECG biometrics and discuss future possibilities.
Curriculum vitæ
João Ribeiro Pinto received his M.Sc. in Bioengineering (field of Biomedical Engineering) from the University of Porto in 2017, with a thesis on the use of the electrocardiogram (ECG) for biometric recognition of vehicle drivers. He is currently pursuing his Ph.D. degree in Electrical and Computer Engineering at the same university, focused on the use of ECG, face, and other data for seamless and continuous biometric recognition and wellbeing monitoring, especially for drivers/passengers inside intelligent vehicles. João has co-authored over thirty publications in biometrics and several other topics related to machine learning and pattern recognition, which have been cited over three hundred times by his peers.
We thank our media partner:
