Work in ML Perception team

Support Perception Algorithm development for autonomous vehicle/ADAS application, which is deployed in R1T/Amazon Delivery Van.

1. Support L2+ ADAS algorithm development, which is deployed in multiple product lines of NA OEM.

2. Develop core algorithm for object tracking (radar and camera based) and perception algorithm/threat assessment in ADAS;

3. Working in the multidisciplinary environment, interacting with multiple teams ranging from software integration, system/tools team globally.

• Post-process Velodyne (VLP-16) sensor data packets in C/C++ and employ DBSCAN algorithm for real time segmentation.

• Employ RANSAC for object detection, global nearest neighbor and kalman filter for data association and object tracking.

• Research on camera techniques, in terms of lane detection and traffic sign classification by deep learning.