Text Driven Temporal Segmentation of Cricket Videos

Abstract

Abstract. In this paper we address the problem of temporal segmentation of videos. We present a multi-modal approach where clues from different information sources are merged to perform the segmentation. Specifically, we segment videos based on textual descriptions or commentaries of the action in the video. Such a parallel information is available for cricket videos, a class of videos where visual feature based (bottom-up) scene segmentation algorithms generally fail, due to lack of visual dissimilarity across space and time. With additional topdown information from textual domain, these ambiguities could be resolved to a large extent. The video is segmented to meaningful entities or scenes, using the scene level descriptions provided by the commentary. These segments can then be automatically annotated with the respective descriptions. This allows for a semantic access and retrieval of video segments, which is difficult to obtain from existing visual feature based approaches. We also present techniques for automatic highlight generation using our scheme.

Cricket Stroke Recognition Using Hard and Soft Assignment Based Bag of Visual Words

“…They use an AlexNet with an SVM for classification. Sankar et al [19] and Sharma et al [23] have focused on the shot segmentation problem in Cricket videos using text alignment.…”

Automatic summarization of broadcast cricket videos

“…KP Sankar et al [5] have proposed a text driven temporal segmentation. They use annotated text data from a website and align the video to it.…”