Sports video annotation is important for sports video semantic analysis such as event detection and personalization.
We propose a novel approach for sports video semantic annotation and personalized retrieval. Different from the state of the art sports video analysis methods which heavily rely on audio/visual features, the proposed approach incorporates web-casting text into sports video analysis.
Compared with previous approaches, the contributions of our approach include the following.
1) The event detection accuracy is significantly improved due to the incorporation of web-casting text analysis.
2) The proposed approach is able to detect exact event boundary and extract event semantics that are very difficult or impossible to be handled by previous approaches.
3) The proposed method is able to create personalized summary from both general and specific point of view related to particular game, event, player or team according to user’s preference.
We present the framework of our approach and details of text analysis, video analysis, text/video alignment, and personalized retrieval. The experimental results on event boundary detection in sports video are encouraging and comparable to the manually selected events.
The evaluation on personalized retrieval is effective in helping meet users’ expectations. A Novel Framework for Semantic Annotation and Personalized Retrieval of Sports Video
One can either watch the full-length sports broadcast recorded on his TiVo box, or watch the summarized highlights from the same broadcaster one day later.
If a sports fan is unable to catch the live action on TV, his/her only means of keeping informed is via multimedia message service (MMS) subscription.
The MMS of the crucial highlights can only be received during the half-time breaks or after the game, since video post-editing is still done manually in the studio. Even if these MMS alerts can be expedited, the mode of transmission is still one-to-many, and clearly unable to meet some viewers’ appetites who may only be interested in the events related to particular player or team.
We present a novel framework for semantic annotation, indexing and retrieval of sports video.
In the framework, we incorporate web-casting text analysis, broadcast sports
Video analysis and text/video alignment to extract event semantics and detect event boundaries, which are important components for sports video annotation, indexing, and retrieval.
In particular, we use soccer and basketball games as our initial target because they are not only globally popular sports but also present many challenges for video analysis due to their loose and dynamic structures compared with other sports such
As tennis.
Since the proposed framework is generic, we believe our approach can be extended to other sports domains.
.