Motovis gathers a world-class team of deep learning and computer vision algorithms and has been holding world leading position in the direction of deep learning and visual SLAM. Motovis confronts with complex automatic driving scenes, adopts the most cutting-edge algorithms and technologies, elaborately designs and optimizes neural networks, achieves multi-angle and multi-target intelligent detection, obstacle avoidance and autonomous positioning of the vehicle, and solves a large number of corner cases troubling the industry to pave the way for the realization of more reliable and safe automatic driving.
Motovis pioneers a deep learning embedded route in the field of autonomous driving. In accordance with standards of the automobile product system, Motovis overcomes the limitation of chips and other engineering challenges and takes the lead to make the highly intelligent computing engine run concurrently and effectively on the embedded processor to deal with various complex visual perception tasks reliably in real-time. The technical difficulty of running massive and complicated intelligent real-time tasks on low-cost and low-power-consumption small on-board hardware systems is overcome.
Motovis has collected mass data which cover the road network systems, road conditions, meteorology and weather changes in 30 provinces and autonomous regions in China. The total collected mileage has exceeded 7 million km, lasting for more than three years. Motovis has established a complete and strict data processing operation procedure independently to classify and process the collected data automatically with high quality, and is promoting the continuous progress of algorithm performance and function with high-quality data.