What you will do:

• Research and develop efficient deep learning models that detect obstacles around the ego-car
• Research and develop algorithms and models to understand the obstacles (e.g., car-door-open) and complex scenes (e.g., traffic light, construction zone, human directing traffic)
• Research and develop algorithms and models to tackle long tail obstacles and scenes (e.g., potholes, trash bags, falling leaves, extreme weather, etc)
• Research and develop algorithms and models to track obstacles and predict their locations, velocities, accelerations, orientations, shapes, etc
• Research and develop algorithms and models to accurately calibrate the sensors and further efficiently verify the correctness of the calibration
• Write onboard code to integrate the algorithms and models into perception computational graph and deploy it to cars

Requirements:

• Have strong abilities in both algorithms and C++ coding; or have strong expertise in at least one of these fields: computer vision, machine learning, and model optimization
• Solid skills in problem solving, algorithms, and data analysis
• Good communication and coordination skills
• Self-motivated and believe in the future of self-driving-cars

Plus:

• Rich experience in C++ development, multithreading, memory management, computing optimization 
• Rich experience in multi-sensor fusion and 3d/2d tracking
• Publications on top-tier conferences and journals