Human tracking with variable prediction steps based on Kullback-Leibler divergence

Authors:
Noriko Takemura;Yutaka Nakamura;Yoshio Matsumoto;Hiroshi Ishiguro
Affiliations:
Department of Adaptive Machine Systems, Graduate School of Engineering, Osaka University, Osaka, Japan 565-0871;Department of Adaptive Machine Systems, Graduate School of Engineering, Osaka University, Osaka, Japan 565-0871;National Institute of Advanced Industrial Science and Technology (AIST), Ibaraki, Japan;Department of Systems Innovation, Graduate School of Engineering Science, Osaka University, Osaka, Japan
Venue:
Artificial Life and Robotics
Year:
2010

Citing 3
Cited 0

Learning Patterns of Activity Using Real-Time Tracking

IEEE Transactions on Pattern Analysis and Machine Intelligence
Memory-Based Attention Control for Activity Recognition at a Subway Station

IEEE MultiMedia
Tracking Multiple Occluding People by Localizing on Multiple Scene Planes

IEEE Transactions on Pattern Analysis and Machine Intelligence

Quantified Score

Hi-index	0.00

Visualization

Abstract

This article deals with a path-planning problem in tracking humans in order to obtain detailed information about human behavior and characteristics. In our method, path planning is performed based on Kullback-Leibler (KL) divergence between the predicted distribution of all human positions and the intensity of the field of view of the agents. The number of steps predicted is determined according to the consistency of the prediction. Experimental results show that when the prediction of human movement is accurate, long-term prediction is useful for path planning. On the other hand, when prediction is inaccurate, long-term prediction might not be useful. Our path-planning method works well even under changing circumstances by changing the length of the predictions.