Neuromorphic Simultaneous Localization and Mapping Using Oculi Sensors

Our world is moving at a fast pace, companies and countries are in the process of developing innovative technologies that would make a real change in all sectors, from autonomous flying drone taxis in Dubai to a variety of artificial intelligence, robots, augmented and virtual reality and self-driving vehicle projects. Conversely, Simultaneous Localization and Mapping, or SLAM, is integrated in one way or another into such projects. SLAM is useful for a variety of indoor, outdoor, aerial, and underwater applications for both manned and automated vehicles. In fact, SLAM is building a high level of intelligence within the robot or any other mobile machinery, which is exactly what the world is trying to adapt to. The proposed project aims at implementing SLAM using a stereo neuromorphic sensor. The suggested algorithm is a combination of both stereo and event-based algorithms. The sensor used in this implementation is the “OCULI SPU ” which is an event-based sensor that mimics biology in speed, parallel processing and selectivity. Finally, the performance of the SPU is compared to the traditional stereo setup when it comes to the implementation of SLAM.

Project Details