Learning Dynamic Event Descriptions in Image Sequences

Automatic detection of dynamic events in video sequences has a variety of applications including visual surveillance and monitoring, video highlight extraction, intelligent transportation systems, video summarization, and many more. The main challenge in most real-world applications is the learning of the event descriptors from limited data in the presence of noise and variations in the occurrence of such events. The main contribution of this work is a semi-supervised learning method in the aforementioned setting for detecting dynamic events in video sequences. Concretely we introduce a stochastic context-free grammar approach for representing the events and learn the event descriptors using an entropy minimization-based semisupervised method. Experimental results demonstrating the efficacy of the learning algorithm and the event detection method applied to real-world video sequences are presented. Keywords Event detection, stochastic context-free grammars, semisupervised learning, en...

• Type: Conference
• Year: 2007
• Where: CVPR
• Authors: Harini Veeraraghavan, Nikolaos Papanikolopoulos, Paul R. Schrater
• Keywords: Computer Vision, CVPR 2007, Event Detection Method, Keywords Event Detection, Real-world Video Sequences, Semi-supervised Learning Method, Stochastic Context-free Grammars