The selected candidate will work closely with top researchers to develop cutting-edge perception algorithms to address real-world problems. The task of perception is to detect and understand the surrounding environment of the autonomous vehicles, to support L2 and L3 automated driving features. A real time virtual representation of the local environment is constructed by processing input data mainly from cameras, together with those from other sensors like radars, IMU and ultrasonic.
Responsibilities:
o Develop deep learning based perception algorithms, including object detection, lane detection, segmentation, depth estimation, and so on.
o Develop state of the art network structures and training schemes, and evaluate their performance on vision problems
o Perform deep network compression based on ASIC platforms
o Keep up with latest technologies in deep learning and computer vision literature and propose innovative ideas
o Publish patent and research papers
Qualifications:
o New graduates are welcome
o MS/PHD in computer science, electrical engineering, or equivalent.
o Fluent in both English and Mandarin
o 2+ years’ experience on computer vision / machine learning development
o Research and development experience in one or more of the following areas:
o Deep network development, in particular with visual data
o Object detection and tracking
o Stereo / optical flow / depth estimation
o Multi-view geometry, 3D computer vision
o Structure from motion
o Network compression and optimization
o Network based sensor fusion
o Mathematical optimization
o Solid programming skills with Python (C++ is a plus)

Please refer the Job description for details