TY - GEN
T1 - Automatic segmentation of TV news into stories using visual and temporal information
AU - Mocanu, Bogdan
AU - Tapu, Ruxandra
AU - Zaharia, Titus
N1 - Publisher Copyright:
© Springer International Publishing AG 2016.
PY - 2016/1/1
Y1 - 2016/1/1
N2 - In this paper we propose a new method for automatic storyboard segmentation of TV news using image retrieval techniques and content manipulation. Our framework performs: shot boundary detection, global key-frame representation, image re-ranking based on neighborhood relations and temporal variance of image locations in order to construct a unimodal cluster for anchor person detection and differentiation. Finally, anchor shots are used to form video scenes. The entire technique is unsupervised being able to learn semantic models and extract natural patterns from the current video data. The experimental evaluation performed on a dataset of 50 videos, totalizing more than 30 h, demonstrates the pertinence of the proposed method, with gains in terms of recall and precision rates with more than 5–7% when compared with state of the art techniques.
AB - In this paper we propose a new method for automatic storyboard segmentation of TV news using image retrieval techniques and content manipulation. Our framework performs: shot boundary detection, global key-frame representation, image re-ranking based on neighborhood relations and temporal variance of image locations in order to construct a unimodal cluster for anchor person detection and differentiation. Finally, anchor shots are used to form video scenes. The entire technique is unsupervised being able to learn semantic models and extract natural patterns from the current video data. The experimental evaluation performed on a dataset of 50 videos, totalizing more than 30 h, demonstrates the pertinence of the proposed method, with gains in terms of recall and precision rates with more than 5–7% when compared with state of the art techniques.
KW - Anchor person extraction
KW - News video story segmentation
KW - Relevant interest points
KW - Temporal and visual constrained clustering
UR - https://www.scopus.com/pages/publications/84994424571
U2 - 10.1007/978-3-319-48680-2_57
DO - 10.1007/978-3-319-48680-2_57
M3 - Conference contribution
AN - SCOPUS:84994424571
SN - 9783319486796
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 648
EP - 660
BT - Advanced Concepts for Intelligent Vision Systems - 17th International Conference, ACIVS 2016, Proceedings
A2 - Distante, Cosimo
A2 - Popescu, Dan
A2 - Scheunders, Paul
A2 - Philips, Wilfried
A2 - Blanc-Talon, Jacques
PB - Springer Verlag
T2 - 17th International Conference on Advanced Concepts for Intelligent Vision Systems, ACIVS 2016
Y2 - 24 October 2016 through 27 October 2016
ER -