Show simple item record

dc.contributor.authorBai, Yancheng
dc.contributor.authorXu, Huijuan
dc.contributor.authorSaenko, Kate
dc.contributor.authorGhanem, Bernard
dc.date.accessioned2018-02-07T07:02:24Z
dc.date.available2018-02-07T07:02:24Z
dc.date.issued2018-01-28
dc.identifier.urihttp://hdl.handle.net/10754/627052
dc.description.abstractActivity detection is a fundamental problem in computer vision. Detecting activities of different temporal scales is particularly challenging. In this paper, we propose the contextual multi-scale region convolutional 3D network (CMS-RC3D) for activity detection. To deal with the inherent temporal scale variability of activity instances, the temporal feature pyramid is used to represent activities of different temporal scales. On each level of the temporal feature pyramid, an activity proposal detector and an activity classifier are learned to detect activities of specific temporal scales. Temporal contextual information is fused into activity classifiers for better recognition. More importantly, the entire model at all levels can be trained end-to-end. Our CMS-RC3D detector can deal with activities at all temporal scale ranges with only a single pass through the backbone network. We test our detector on two public activity detection benchmarks, THUMOS14 and ActivityNet. Extensive experiments show that the proposed CMS-RC3D detector outperforms state-of-the-art methods on THUMOS14 by a substantial margin and achieves comparable results on ActivityNet despite using a shallow feature extractor.
dc.publisherarXiv
dc.relation.urlhttp://arxiv.org/abs/1801.09184v1
dc.relation.urlhttp://arxiv.org/pdf/1801.09184v1
dc.rightsArchived with thanks to arXiv
dc.titleContextual Multi-Scale Region Convolutional 3D Network for Activity Detection
dc.typePreprint
dc.contributor.departmentVisual Computing Center (VCC)
dc.contributor.departmentComputer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division
dc.contributor.departmentElectrical Engineering Program
dc.eprint.versionPre-print
dc.contributor.institutionBoston University, USA
dc.identifier.arxividarXiv:1801.09184
kaust.personBai, Yancheng
kaust.personGhanem, Bernard
refterms.dateFOA2018-06-14T04:57:24Z


Files in this item

Thumbnail
Name:
1801.09184v1.pdf
Size:
2.821Mb
Format:
PDF
Description:
Preprint

This item appears in the following Collection(s)

Show simple item record