30 Billion Images a Second: Tomorrow’s AI Cities Focus on Intelligent Video Analytics

Thirty billion images a second. One hundred trillion images an hour.

That's how much content will be captured by 2020 by surveillance cameras across the globe. These one billion cameras - twice today's number - will be at traffic intersections, transit stations and other public areas, helping to make our cities safer and smarter. They'll be in retail stores, service centers, warehouses, and more, gathering information to boost sales, track inventory, improve service.

But how to make sense of this staggering number of pixels?

Traditional methods of video processing - using human monitoring or hand-crafted computer vision algorithms - aren't up to the task. But a branch of artificial intelligence called deep learning offers a powerful, scalable method for extracting the most out of these vision systems.

Deep learning-powered AI computers will be able to watch and listen - and understand what's on screen - at phenomenal speeds. They'll pull together massive datasets from disjointed locations, transforming pixels into insight.

The result will be AI Cities. We'll be able to recognize when there's an accident, a pet or child is lost, or someone is experiencing harm. An AI City will be a thinking robot, with many eyes, programmed to help keep people safe.

At GTC China last week, NVIDIA CEO Jen-Hsun Huang described how AI computing will help us gain meaningful insights from this flood of information reliably and accurately. He also announced our first AI City partners - China's HikVision, Dahua and UniView.

These companies are the top three market share leaders for video surveillance products and, according to analyst firm IHS, make up more than half of the physical security equipment market in China and a quarter globally.

This momentum continued today at GTC Taiwan, where we just announced SenseTime - a Hong Kong based leader in facial and object recognition, image search, and surveillance - as an AI City partner.

Our work with companies like these fundamentally changes how video is captured, stored and analyzed. Using technologies such as our NVIDIA Jetson TX1, they can easily add deep learning capabilities at very low power. Features such as multi-object classification, facial recognition and behavior analysis will be able to operate on multiple channels in real time. And we feel this is just the start of something truly visionary.

If you missed GTC China and Taiwan, be sure to join us at the Securing New Ground Conference on Oct. 19, in New York, where Deepu Talla, general manager and vice president of our Tegra business unit, will tackle the topic of security and deep learning.