AI dashcam image retrieval method improves urban localization accuracy
Leveraging billions of images in the company’s connected vehicle network, the new visual localization approach, says the company, greatly improves GPS location accuracy in urban areas and offers a scalable solution for navigation and autonomous vehicle applications. As a benchmark to evaluate the method, the company is also releasing a dataset and benchmark based on anonymized dash cam and GPS information from its network to advance the research of visual localization for safety applications.
“This new localization method makes it possible for Nexar to deliver on our founding promise, which is to help rid the world of collisions,” says Nexar co-founder and Chief Technology Officer Bruno Fernandez-Ruiz. “And the benefits will go far beyond our network – this approach could one day allow autonomous vehicles to reliably navigate cities.”
“It’s just as accurate and far less expensive than structure-based techniques such as lidar, which are limited in scale and expensive to compute,” says Fernandez-Ruiz. “So the potential is really tremendous.”
By pairing Nexar-powered dash cameras with the the company’s app, drivers join Nexar’s global safe-driving network, where every vehicle is alerted to what’s happening on the road ahead with the help of other vehicles around it. To deliver these critical alerts, says the company, it needs an efficient and accurate way of knowing in real time exactly where vehicles are on the road – a challenge for solutions like GPS in dense urban environments.
The AI-powered image retrieval algorithm, says the company, promises to dramatically improve localization in cities, solving a problem that has long-plagued both rideshare operators and navigation apps, as well as autonomous vehicle manufacturers. Its research of crowd-sourced data of over 250,000 driving hours in New York City found that at least 40% of rides suffered GPS errors of 10 meters or more due to the “urban canyon” effect.
Using a hybrid coarse-to-fine approach that leverages visual and GPS location cues, the new method for localization is based on the company’s continuously growing database of fresh road imagery observed by its network, which is currently indexing over 10 billion new road images per month. The company says its method solves for the challenges presented by other mapping companies, such as Google, which can use a vast Street View image repository, but lack the minute-level freshness that Nexar provides.
To this end, the company has trained a deep learning model to identify a driver’s accurate location using its massive archive of anonymized images. The archive includes billions of images from more than 400 million miles driven on the Nexar network.
The company says that experiments confirm that this localization approach is highly effective in challenging urban environments, reducing the distribution of localization errors by an order of magnitude. This method will be used to deliver alerts to Nexar users, including dangerous intersections and collisions on the road ahead.
For more, see the company’s research paper on the approach: “Accurate Visual Localization for Automotive Applications.”
Related articles:
Smart ADAS device detects, predicts potholes
AI helps vehicles to find their way through roadworks
Ambarella, Momenta collaborate on autonomous HD mapping platform
Street-level analytics startup accelerates 3D reality capture plans
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :
