中文
Home / News & Events / Events / 正文

Human Motion Recognition Methods Using Radars

Release time:November 27, 2018

Topic:Human Motion Recognition Methods Using Radars

Speaker:Branka Jokanovic, SeniorEngineer, Aptiv PLC

Time:10:00 AM, December 5

Venue:F708, New Main Building

Abstract:

“A pedestrian is killed by a car roughly every 90 minutes in the United States. And until this week, all of those drivers were human.” This was a newspaper headline in March 2018 after a pedestrian was struck and killed by a self-driving car. Even though self-driving vehicles do not pose any more of a risk to pedestrians than cars piloted by humans, making autonomous vehicles drive safely in dense urban environment is still a challenge. This talk addresses how a radar system can recognize different human motions, including a pedestrian walking.

Radar is a remote sensor that has been proven successful in human motion recognition. Compared with LiDAR and camera, radar provides reliable monitoring while being robust to lighting, temperature and weather conditions. This presentation covers traditional and novel recognition methods of human motion radar signatures. Traditional methods are based on manual feature extraction or Principal component analysis (PCA). Novel methods are based on deep learning techniques. Deep learning has emerged as the key part in the field of artificial intelligence due to its powerful brain-mimicking neural network structures. These complex structures allow an automated way of learning and capturing the intricate properties of the human motion signatures in different domains. Experimental results demonstrate that the deep learning methods provide high accuracy compared with the traditional methods.

Biography of the Speaker:

Branka Jokanovic is a radar algorithm development engineer at Aptiv PLC. After completing her M.Sc. degree at University of Montenegro in 2012, she joined Villanova University where she received her Ph.D. degree in 2017. During her graduate studies, her research focused on time-frequency analysis, compressive sensing, deep learning and their application to human motion recognition using radars. Since joining Aptiv, she has been working on advanced MIMO automotive radar systems. Her current focus is on pedestrian and other vulnerable road objects detection.

 

School of Electronic and Information Engineering