Ambarella, Inc., which develops low-power, HD and Ultra HD video processing semiconductors, is demonstrating its fully autonomous Embedded Vehicle Autonomy (EVA) vehicle on the roads of Silicon Valley to industry analysts and customers.

Ambarella says that EVA has been trained to handle different traffic scenarios presented by Silicon Valley’s challenging urban environment.

EVA combines software and algorithms based on more than 20 years of autonomous vehicle research with Ambarella’s “low-power CV1 embedded computer vision processors” based on its CVflow architecture.

EVA’s high-resolution stereovision cameras deliver the “360-degree short and long distance viewing capability” needed for advanced perception and precise self-location. EVA also includes “sensor fusion of the vision information with Radar and map data,” to provide the necessary information for path planning and merging maneuvers without the need for more lidar systems.

“High resolution 8-Megapixel stereovision combined with superior perception in challenging lighting conditions allows EVA to “see” its surroundings with much higher reliability than was previously possible,” explains Professor Alberto Broggi, General Manager of Ambarella Italy.

“Moving to an implementation based on dedicated Ambarella CVflow processors brings us much closer to making self-driving cars a practical reality.” 

According to Ambarella, EVA’s CV1-based stereovision cameras provide a perception range of over 150 meters for stereo obstacle detection, and over 180 meters for monocular classification.

Stereovision processing enables “detection of generic obstacles” without training, which allows for more robust decisions to be made, Ambarella says.  

EVA also uses stereovision to recognize visual landmarks, and the autonomous vehicle uses HD map information for high precision localization, even in instances when the GPS signal is weak or not available, such as in dense urban locations.

EVA has a variety of unique features, including but not limited to, automatic calibration, traffic light detection, and CNN classification for vehicle, pedestrian, and bicycle/motorcycle.