Event detection with zero example: Select the right and suppress the wrong concepts

Event detection with zero example: Select the right and suppress the wrong concepts

Lu, Y.J.
Zhang, H.
de Boer, M.H.T.
Ngo, C.W.

2016

Complex video event detection without visual examples is a very challenging issue in multimedia retrieval. We present a state-of-the-art framework for event search without any need of exemplar videos and textual metadata in search corpus. To perform event search given only query words, the core of our framework is a large, pre-built bank of concept de- Tectors which can understand the content of a video in the perspective of object, scene, action and activity concepts. Leveraging such knowledge can effectively narrow the se- mantic gap between textual query and the visual content of videos. Besides the large concept bank, this paper focuses on two challenges that largely affect the retrieval performance when the size of the concept bank increases: (1) How to choose the right concepts in the concept bank to accurately represent the query; (2) if noisy concepts are inevitably cho- sen, how to minimize their inuence. We share our novel insights on these particular problems, which paves the way for a practical system that achieves the best performance in NIST TRECVID 2015. © 2016 ACM.

ICT
DSC - Data Science
TS - Technical Sciences
0Ex
Concept bank
Concept selection
Multimedia event detection
Semantic pooling
Video search

http://resolver.tudelft.nl/uuid:e4740464-0ae8-44b7-87d5-8d742af7009a

https://doi.org/10.1145/2911996.2912015

745667

ACM

9781450343596

Proceedings of the 2016 ACM International Conference on Multimedia Retrieval ICMR’16 June 6–9, 2016, New York, NY, USA, 127-134

conference paper