With the improvement in Artificial Intelligence, new deep neural network architectures can now achieve face recognition accuracy similar to a human level. One drawback of this approach is that, it is prone to face spoofing attacks where an impostor can gain access to the system by presenting a copy of the image to the camera. In this talk, Dr. Ausif Mahmood will describe the state-of-the-art approaches to face recognition using deep neural networks. Dr. Ausif Mahmood also will present deep architectures for face liveness detection that use a combination of texture analysis and convolutional neural network (CNN) to classify the captured image as real or fake. Furthermore, Dr. Ausif Mahmood will present insight into the enhancement of the face liveness detection architecture by evaluating different deep architectures, which include deep residual network, and the inception network version 4. Dr. Ausif Mahmood will evaluate the performance of each of these architectures on the NUAA dataset, and will present experimental results showing under what conditions an architecture would be better suited for face liveness detection. While the residual network gave competitive results, the inception network version 4 gave the highest accuracy of 100% in the detection of liveness of images, outperforming all previously proposed state-of-the-art methods.

Dr.  Mahmood is the Director of The School of Engineering at the University of Bridgeport in Bridgeport, CT. His research areas involve Artificial Intelligence, Computer Vision, Machine Learning, and Deep Learning - bridging ties between Computer Science and Data Science.

This presentation will take place virtually on September 17th at 1:00pm - 2:00pm.  Registrants will be provided video conferencing information including a Zoom link a day prior to the event..