I have a visceral memory of sitting around a roaring campfire one night in Yosemite years ago. The memory is vivid because at a certain point, I turned away from the campfire and looked into the surrounding darkness only to see the eyes of a pack of coyotes patiently watching from every direction. This experience has given me a rather unique perspective on edge-based surveillance.
Without a doubt, one of the most compelling applications of deep learning is coyote, I mean, computer vision. Computer vision has become so pervasive that we experience it without even knowing it. For example, every time we pick up a smartphone it effortlessly unlocks with facial recognition, no more typing in a PIN or fumbling with a fingerprint. The truth is that deep learning is pushing the limits of what is possible in Computer Vision (CV). With methods such as Convolutional Neural Networks (CNNs), problems thought to be unsolvable—such as image classification—are now being solved with super-human accuracy. Many CV techniques invented in the past 20 years, including traditional feature-based approaches, have become irrelevant because of DL. Advancement in DL and device capabilities, including processors, memory capacity, power consumption, image sensor resolution, and optics, are enabling new computer vision-based applications at the edge.
But it can feel lonely sometimes operating at the leading edge of advanced technology. As product creators seek to incorporate vision capabilities into edge AI hardware and software, where can they turn to find inspiration and practical information to guide them? The answer is the Edge AI and Vision Alliance. Anyone designing practical computer vision solutions, including the selection of chips, cameras, algorithms, tools, and programming languages, should check out the Edge AI and Vision Alliance. Their membership meetings and events like the Embedded Vision Summit provide practical know-how for system designers looking to effectively incorporate AI technologies into their designs.
As a member of the alliance, Expedera gains numerous advantages, including networking with the largest number of potential AI development partners and customers (more than 100 at current count) in one setting. Additionally, Expedera will be exhibiting at 2022 Embedded Vision Summit, happening May 17-19th, just a mile or so away from Expedera HQ in Santa Clara. Perhaps most exciting, besides membership and exhibiting, Expedera will be presenting our Origin™ technology as part of the conference’s Enabling Technologies Track. More information on this to following the coming months.
Have questions? We’d love to hear from you. Drop us a note here, and we will be in touch.