Zongwen Mu

Zongwen Mu has three years of work experience, beginning in 2018. In 2018, they worked as a Navigation and Test Engineer at Zhejiang Yonda Technology Co. Ltd. In 2020, they were a Research Intern at the Computational Engineering and Robotics Lab at Carnegie Mellon University. In 2021, they have been a Perception Engineer at AutoX.ai. At AutoX.ai, they developed a time-sequential deep learning model for traffic light classification using ResNet as the backbone and Transformer with auxiliary supervision as the neck, achieving a 0.81 mAP on a customized dataset. Zongwen also designed a clip sampler to generate sequence labels from single frames dataset and used multi-processing to accelerate the processing time from over 400s to 90. At the Computational Engineering and Robotics Lab, they utilized YOLO detector on monocular camera videos to detect dynamic obstacles in tunnel, generated training data for engineering vehicles in simulation, and transformed training data into COCO format. Zongwen also trained neural networks on a customized dataset and enabled the detector to detect excavators and bulldozers correctly.

Zongwen Mu obtained a Bachelor's degree in Mechatronic Engineering from Zhejiang University in 2015-2019. Zongwen then pursued a Master of Science in Mechanical Engineering from Carnegie Mellon University in 2019-2020, and obtained a Master of Science certification from Carnegie Mellon University in December 2020.