In this talk, I will investigate breakthroughs in NLP and examine platforms that are widely used today as end product NLP solutions. As the world of AI, more particularly Deep Machine Learning, experiences a renaissance, one of the critical, more outspoken issues, revolves around the differential performance of the black box technology. In this work, I will extend the lens of bias and differential performance to examine NLP solutions from major vendors. Further, I will illustrate that these NLP black boxes still contain “bias.”
Dr. Karl Ricanek Jr. received his Ph.D. in Electrical Engineering from North Carolina A&T State Univ in 1999. His 1999 dissertation focused on the development of an efficient compact face representation that used an artificial neural network to solve the problem of face recognition across poses. Over the next 30 years, Dr. Ricanek made significant advancements in biometrics, but most prominently in the face space. He developed the field of facial analytics that was an outgrowth of a seminal article in Computer 2010; made deep contributions in face-based age progression algorithms and automatic facial landmarking and helped to establish the domain of periocular recognition and eyebrow recognition. He holds two patents in facial analytics. He is a member of multiple international scientific working groups for face recognition, biometrics, and machine learning where he shares his insights with governments and industries around the world.
Prior to starting the doctorate program, Dr. Ricanek was an engineer with the Naval Undersea Warfare Center (NUWC) where he worked on expert systems, combat control systems, and other platforms for the U.S. Nuclear Submarine Fleet. He developed the first Object-Oriented Window Environment for the Combat Systems Evaluation and Analysis Lab. He later implemented the first Digital Nautical Charts platform that leveraged ocean-based GIS solutions. He was awarded multiple awards and honorifics for his service with NUWC.
Upon completion of his doctorate, Dr. Ricanek joined Corning Optical Fiber where he developed intelligent manufacturing systems. While with Corning he led multiple international engineering projects and teams, developed machine vision solutions to improve optical fiber manufacturing processes, and worked on technology for computed tomography and diffraction-based metrology.
Dr. Ricanek joined the University of North Carolina, Wilmington, in 2003 where he focused on advanced biometric and computer vision research for the Intelligence Community (IC). He has been the lead PI or co-PI on more than 40 grants and contracts with the IC and Department of Defense. Dr. Ricanek has worked with more than 100 undergraduate and graduate students at UNCW and across the world.
Dr. Ricanek has more than 85 peer-reviewed articles and book chapters in computer vision, biometrics, facial analytics, craniofacial morphology, and artificial intelligence. He has commercialized his patents; licensed his technology and products around the globe and has spun out three companies. Further, Dr. Ricanek developed the first viral web app Face My Age in the summer of 2014 that launched many of the AI face apps that you see today.