TY - JOUR
T1 - Unsupervised object-level video summarization with online motion auto-encoder
AU - Zhang, Yujia
AU - Liang, Xiaodan
AU - Zhang, Dingwen
AU - Tan, Min
AU - Xing, Eric P.
N1 - Publisher Copyright:
© 2018 Elsevier B.V.
PY - 2020/2
Y1 - 2020/2
N2 - Unsupervised video summarization plays an important role on digesting, browsing, and searching the ever-growing videos every day, and the underlying fine-grained semantic and motion information (i.e., objects of interest and their key motions) in online videos has been barely touched. In this paper, we investigate a pioneer research direction towards the fine-grained unsupervised object-level video summarization. It can be distinguished from existing pipelines in two aspects: extracting key motions of participated objects, and learning to summarize in an unsupervised and online manner. To achieve this goal, we propose a novel online motion Auto-Encoder (online motion-AE) framework that functions on the super-segmented object motion clips. Comprehensive experiments on a newly-collected surveillance dataset and public datasets have demonstrated the effectiveness of our proposed method.
AB - Unsupervised video summarization plays an important role on digesting, browsing, and searching the ever-growing videos every day, and the underlying fine-grained semantic and motion information (i.e., objects of interest and their key motions) in online videos has been barely touched. In this paper, we investigate a pioneer research direction towards the fine-grained unsupervised object-level video summarization. It can be distinguished from existing pipelines in two aspects: extracting key motions of participated objects, and learning to summarize in an unsupervised and online manner. To achieve this goal, we propose a novel online motion Auto-Encoder (online motion-AE) framework that functions on the super-segmented object motion clips. Comprehensive experiments on a newly-collected surveillance dataset and public datasets have demonstrated the effectiveness of our proposed method.
KW - Object-level video summarization
KW - Online motion auto-encoder
KW - Stacked sparse LSTM auto-encoder
UR - https://www.scopus.com/pages/publications/85050966685
U2 - 10.1016/j.patrec.2018.07.030
DO - 10.1016/j.patrec.2018.07.030
M3 - 文章
AN - SCOPUS:85050966685
SN - 0167-8655
VL - 130
SP - 376
EP - 385
JO - Pattern Recognition Letters
JF - Pattern Recognition Letters
ER -