AI-generated depth mapping research project by Immervision and Université Laval

  • November 22, 2018

In keeping with its mission to design the most advanced vision systems to bring humanity to any device, Immervision, in partnership with Université Laval, is launching a joint research project with the goal of using artificial intelligence deep learning algorithms to enable 3D depth sensing from a single wide-angle camera.

Super-wide and ultra-wide cameras capture more scene information than traditional narrow-angle ones.  In combination with data from other sensors, this rich vision enables machines to perceive the world with greater detail than humans.  Applications for this super-human perception are virtually limitless and include smartphones, autonomous vehicles, driver assistance systems, robotics, IoT, SLAM, and AR/XR, as well as future appliances and devices.

Dr. Jean-François Lalonde, Associate Professor, Vision and Digital Systems Lab, Université Laval, says « We will research new deep learning algorithms that generate rich depth information from a single camera view.  Working with the combination of wide-angle, high pixel density images captured with Immervision panomorph technology and other sensor data, we aim to take an important step towards full machine perception.”

Patrice Roulet, Immervision Vice-President, Technology and Co-Founder puts it this way: “Even with one of your eyes closed, your other senses and your brain work together so you can understand depth and your environment with just half your usual vision. At Immervision, we continue to research new technology to bring humanity to machines, so they can see, navigate and understand the world like never before.”


Contact:
Immervision

Angus Mackay
+1 514 985-4007 ext. 3037
[email protected]