There is tremendous potential for facial recognition technology, such as informing visually impaired persons if someone they know is approaching them.
I find it difficult to believe just how fast things are moving with regard to using artificial neural networks (ANNs) and deep learning techniques (for example, see Deep learning machine vision system aids blind and visually impaired, Deep learning hits a sweet note, Machine learning platform speeds optimization of vision systems, Unlocking the power of AI for all developers, and Push-button generation of deep neural networks).
Of course, one really interesting application is to perform object detection and identification, including the really tricky task of recognizing and identifying faces in images and videos.
This sort of task benefits from the extreme parallelism offered by FPGAs. Of particular interest are Intel's current generation of FPGAs, whose hard-core DSP slices offer both fixed-point and floating-point capabilities, making them suitable for a wide range of artificial intelligence (AI) and embedded vision applications.
Intel's midrange Arria 10 FPGAs, for example, provide up to 1.5 teraflops (TFLOPs) of single-precision floating-point processing performance, 1.15 million logic elements, and more than a terabit-per-second high-speed connectivity.
Working with Intel's Programmable Solutions Group, engineers at ZTE -- a Chinese multinational telecommunications equipment and systems company headquartered in Shenzhen, Guangdong -- have used Arria 10 FPGAs to implement a cloud inferencing application using a convolutional neural network (CNN) algorithm to perform high-accuracy facial recognition on more than 1,000 images/frames per second.
(Source: Getty Images)
Traditionally, facial recognition technology has been primarily of interest to law enforcement agencies, who use these systems to isolate and identify faces in crowds (at airports, for example). But there is tremendous potential for this technology in a wide variety of applications, such as informing blind and visually impaired persons if someone they know is approaching them.
Facial recognition to replace passports in Australian airports, but at what ethical cost?
Troubling study claims machine-learning algorithm predicts criminal tendency based on facial traits
Omron Face Detection Technology Fact Sheet
In the case of your humble narrator, I'm dreadfully bad at remembering names and recognizing faces. This problem is exacerbated by the fact that I attend a lot of conferences, and people are always coming up to me saying things like "Max, how are you?" And I have to dance around (conversationally-speaking), desperately trying to work out who they are and how I know them.
Based on this, I can easily envision how useful a Google Glass-type device would be if it had the ability to identify people in a crowd and present me with useful information -- possibly using bone conduction so as to leave my ears free and also prevent anyone else knowing what was going on -- like "Professor Cuthbert Dribble is approaching. You last saw him at the Embedded Conference in Copenhagen in November 2015. He loves cats and it's his turn to buy a round of beers."
How about you? What applications can you see (no pun intended) for this sort of facial identification and recognition technology?
— Max Maxfield, Editor of All Things Fun & Interesting