报告题目：Deep Learning for Autonomous Vehicle Perception

报 告 人：Xinming Huang教授Worcester Polytechnic Institute

报告时间：2019年8月6日上午10：30

报告地点：南一楼中311室

摘要:Since DARPA introduced the autonomous ground vehicle grand challenge in 2007, research on autonomous vehicle has grown rapidly in both academia and industry. The core technologies include computer vision, machine learning, sensor fusion, autonomous control, vehicular communications, and etc. The emergence of autonomous vehicles on roads also requires attentions on many related issues such as transportation infrastructure, legal policy, and insurance liability. This talk presents the current research on autonomous vehicles at WPI. The main focus will be perception and machine learning algorithms and their real-time implementations on software/hardware platforms. The vision and LiDAR based perceptions include road segmentation, lane detection, traffic sign detection and classification, traffic light detection and recognition, road marking detection, vehicle detection, pedestrian detection, and etc. Deep learning has been applied by training convolutional neural network models with sensor data. WPI research team have built a fully functional autonomous vehicle prototype and collected a large amount of data using on-board sensors, such as cameras, LiDAR, radars, GPS and IMU. In addition, both GPU and FPGA-based hardware platforms have been employed as accelerators for real-time processing. Our research goal is to propose new algorithms and techniques for automated driving and evaluate them on road testing.

简历：Xinming Huang is the Joseph Samuel Satin distinguished professor in the Department of Electrical and Computer Engineering at Worcester Polytechnic Institute (WPI). Dr. Huang received his PhD in electrical engineering from Virginia Tech in 2001. He was a member of technical staff with the Bell Labs of Lucent Technologies before he joined the faculty at WPI in 2006. He has received numerous recognitions including DARPA young faculty award, IBM faculty award, Bell Labs annual excellence award, and WPI’s dean’s excellence professor award. He has published more than 140 referred papers in IEEE journals and conferences. He is an Associate Editor for IEEE Transactions on Signal Processing. His current research interests are focused on deep learning hardware architecture for autonomous vehicles, smart health, and internet of things.