Video summarization guiding evaluative rectification for industrial activity recognition

Voulodimos, AS; Doulamis, AD; Kosmopoulos, DI; Varvarigou, TA

dc.contributor.author	Voulodimos, AS	en
dc.contributor.author	Doulamis, AD	en
dc.contributor.author	Kosmopoulos, DI	en
dc.contributor.author	Varvarigou, TA	en
dc.date.accessioned	2014-03-01T02:53:31Z
dc.date.available	2014-03-01T02:53:31Z
dc.date.issued	2011	en
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/36382
dc.subject.other	Accuracy rate	en
dc.subject.other	Activity recognition	en
dc.subject.other	Classification results	en
dc.subject.other	Computational costs	en
dc.subject.other	Data sets	en
dc.subject.other	Expert users	en
dc.subject.other	Industrial activities	en
dc.subject.other	Industrial surveillance	en
dc.subject.other	Key-frames	en
dc.subject.other	Representative sample	en
dc.subject.other	Semantic content	en
dc.subject.other	Training sample	en
dc.subject.other	Training sets	en
dc.subject.other	Video summarization	en
dc.subject.other	Experiments	en
dc.subject.other	Image processing	en
dc.subject.other	Natural language processing systems	en
dc.subject.other	Sampling	en
dc.subject.other	Security systems	en
dc.subject.other	Semantics	en
dc.subject.other	Video recording	en
dc.subject.other	Industry	en
dc.title	Video summarization guiding evaluative rectification for industrial activity recognition	en
heal.type	conferenceItem	en
heal.identifier.primary	10.1109/ICCVW.2011.6130354	en
heal.identifier.secondary	http://dx.doi.org/10.1109/ICCVW.2011.6130354	en
heal.identifier.secondary	6130354	en
heal.publicationDate	2011	en
heal.abstract	In this paper we present a video summarization method that extracts key-frames from industrial surveillance videos, thus dramatically reducing the number of frames without significant loss of semantic content. We propose to use the produced summaries as training set for neural network based Evaluative Rectification. Evaluative Rectification is a method that exploits an expert user's feedback regarding the correctness of an activity recognition framework on part of the data in order to enhance future classification results. The size of the training sample set usually depends on the topology of the network and on the complexity of the environment and activities observed. However, as is shown by the experiments conducted in a real-world industrial activity recognition dataset, using a much smaller but representative sample stemming from our summarization technique leads to significantly higher accuracy rates than those attained by a same size but randomly chosen set. To obtain comparable improvement in accuracy without the summarization technique, the experiments show that a far larger training sample set is needed, therefore requiring significantly increased human resources and computational cost. © 2011 IEEE.	en
heal.journalName	Proceedings of the IEEE International Conference on Computer Vision	en
dc.identifier.doi	10.1109/ICCVW.2011.6130354	en
dc.identifier.spage	950	en
dc.identifier.epage	957	en