Generic architecture for event detection in broadcast sports video
Proceedings of the 3rd international workshop on Automated information extraction in media production
Hi-index | 0.00 |
We present a novel learning-based framework for detecting interesting events in soccer videos. The input to the systemis a raw soccer video. We have learning at three levels-learning to detect interesting low-level features from imageand video data using Support Vector Machines (hereafter, SVMs), and a hierarchical Conditional Random Field- (hereafter, CRF-) based methodology to learn the dependencies of mid-level features and their relation with the lowlevelfeatures, and high level decisions (‘interesting events’) and their relation with the mid-level features: all on the basisof training video data. Descriptors are spatio-temporal in nature - they can be associated with a region in an imageor a set of frames. Temporal patterns of descriptors characterise an event. We apply this framework to parse soccer videos into Interesting (a goal or a goal miss) and Non-Interesting videos. We present results of numerous experiments in support of the proposed strategy.