A neural network containing annotated visual data collected for over 4 years made available for testing autonomous mobility applications by engineers, worldwide
San Jose, CA – January 5, 2023 - Imagry, a developer of mapless autonomous driving technology, announced today that it has unlocked its legacy data set of annotated continuous VIS (visible imaging sensor) camera images for the benefit of engineering and machine learning students and entrepreneurs who want to test their new autonomous mobility applications against real-life scenarios. Any interested party will be able to download full sequence, time-based road perception data that has been collected over the past four years and annotated to provide top-down maps for a birds-eye view. The data can be used to test constructs for artificial intelligence applications, such as autonomous driving and robots.
Imagry Uses a Location-Independent Autonomous Driving Model
Imagry software combines cutting-edge technology that captures and recognizes a real-time video feed with a deep neural network to make complex driving decisions instantly, based on current driving conditions. For the past four years, Imagry has been operating vehicles equipped with its autonomous driving software on public roads in the United States, Germany, and Israel, generating an extensive visual library of driving situations. This library serves as the basis for the Imagry platform which is slated to propel autonomous buses on public roads later this year.
Data Set is a Treasure Trove for Engineers Developing New Mobility Applications
“As veterans in the development of an autonomous mobility platform, we know how important it is to test new applications against a solid data set,” states Mr. Majed Jubeh, Co-founder and Chief Delivery Officer at Imagry. “We’re happy to share our extensive library of annotated images to encourage young engineers and allow them to test their ideas against real-life scenarios.”
Availability
All the necessary code and documentation for loading, parsing, and accessing the visual images and mapping will be housed on a GitHub site. The location of the site will be made available to anyone who registers using this link. Those choosing to download the data will be encouraged to share the applications they develop with the Imagry team to get feedback about their projects.
Imagry is exhibiting at CES 2023, in booth #5874 of the LVCC West Hall.
About Imagry
Imagry (www.imagry.co/) was founded in 2015 by Mr. Majed Jubeh and Mr. Adham Ghazali. The company develops unique mapless driving software for Level 3-4 autonomous vehicles. Imagry's bio-inspired technology collects and processes peripheral information from cameras installed on the vehicle. This data is sent to a computer in the vehicle whose purpose is to perform, in real-time, the actions that will allow the vehicle to drive autonomously (like a human driver). Imagry's unique technology enables the rapid deployment of autonomous vehicles on unknown roads for non-stop journeys. It uses low-cost sensors and eliminates the need for external digital mapping of roads and cities.
###