Robust highlight extraction using multi-stream hidden markov models for baseball video (2005) [1 citations — 0 self]
Abstract:
This paper proposes a robust statistical framework to extract highlights from a baseball broadcast video. We applied multistream Hidden Markov Models (HMMs) to control the weights among different features. To achieve robustness against new highlights, we used a common simple structure for all the HMMs. In addition, scene segmentation and unsupervised adaptation were applied to achieve more robustness against the differences of environmental conditions among games. The precision rate of highlight extracting experiments for eight kinds of highlights from 4.5 hours of digest data was 77.4 % and was increased to 78.7 % by applying scene segmentation. Futhermore, the unsupervised adaptation method improved precision by 2.7 points to 81.4%. These results confirm the effectiveness of our framework. 1
Citations
| 78 | Maximum likelihood linear regression for speaker adaptation of continuous density hidden Markov models – Legetter, Woodland - 1995 |
| 39 | Content-based video indexing of TV broadcast news using Hidden Markov Models – Eickeler, Müller - 1999 |
| 39 | Structure Analysis of Sports Video Using Domain Models – Zhong, Chang - 2001 |
| 29 | Content Analysis of Video Using Principal Components – Sahouria, Zakhor - 1999 |
| 21 | Extract highlights from baseball game video with hidden markov models – Chang, Han, et al. - 2002 |
| 19 | The HTK Book (for HTK version 3.2 – Young, Evermann, et al. - 2002 |
| 9 | Event Detection and Summarization in Sports Video – Li, Sezan - 2001 |
