A feature-based algorithm for detecting and classifying scene breaks
Proceedings of the third ACM international conference on Multimedia
Segmentation of Lecture Videos Based on Text: A Method Combining Multiple Linguistic Features
HICSS '04 Proceedings of the Proceedings of the 37th Annual Hawaii International Conference on System Sciences (HICSS'04) - Track 1 - Volume 1
Video Shot Detection Using Hidden Markov Models with Complementary Features
ICICIC '06 Proceedings of the First International Conference on Innovative Computing, Information and Control - Volume 3
Optimal Shot Boundary Detection Based on Robust Statistical Models
ICMCS '99 Proceedings of the 1999 IEEE International Conference on Multimedia Computing and Systems - Volume 02
Parsing news video using integrated audio-video features
PReMI'05 Proceedings of the First international conference on Pattern Recognition and Machine Intelligence
Decentralized quickest change detection
IEEE Transactions on Information Theory
Shot-boundary detection: unraveled and resolved?
IEEE Transactions on Circuits and Systems for Video Technology
| Hi-index | 0.00 |
Temporal segmentation of a video into its constituent shots is the basic step towards the exploration about the organization of digital video for all higher level analysis. Video shot detection methods in the literature mostly involve heuristics and fail to perform satisfactorily under varied shot detection scenarios. Though model based shot recognition methods are popular, they are inadequate when a given test video sequence contains transitions. Not much work has been reported which deal with the changes in activities in areas where we have to recognize the activities over a long video sequence. We formulate this as a novel N-class, model based shot detection problem and present a stochastic, asymptotically optimal procedure as a solution to such a scenario, so that neither changes in content nor the types of shot transition hinder the decision making process. A hidden Markov model (HMM), trained using a few relevant features from the different classes of frame sequences is employed to achieve this goal. We present extensive experimental results to demonstrate the effectiveness of our method.