File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Conference Paper: Pedestrian parsing via deep decompositional network

TitlePedestrian parsing via deep decompositional network
Authors
Keywordspedestrian parsing
deep learning
Issue Date2013
Citation
Proceedings of the IEEE International Conference on Computer Vision, 2013, p. 2648-2655 How to Cite?
AbstractWe propose a new Deep Decompositional Network (DDN) for parsing pedestrian images into semantic regions, such as hair, head, body, arms, and legs, where the pedestrians can be heavily occluded. Unlike existing methods based on template matching or Bayesian inference, our approach directly maps low-level visual features to the label maps of body parts with DDN, which is able to accurately estimate complex pose variations with good robustness to occlusions and background clutters. DDN jointly estimates occluded regions and segments body parts by stacking three types of hidden layers: occlusion estimation layers, completion layers, and decomposition layers. The occlusion estimation layers estimate a binary mask, indicating which part of a pedestrian is invisible. The completion layers synthesize low-level features of the invisible part from the original features and the occlusion mask. The decomposition layers directly transform the synthesized visual features to label maps. We devise a new strategy to pre-train these hidden layers, and then fine-tune the entire network using the stochastic gradient descent. Experimental results show that our approach achieves better segmentation accuracy than the state-of-the-art methods on pedestrian images with or without occlusions. Another important contribution of this paper is that it provides a large scale benchmark human parsing dataset that includes 3,673 annotated samples collected from 171 surveillance videos. It is 20 times larger than existing public datasets. © 2013 IEEE.
Persistent Identifierhttp://hdl.handle.net/10722/273661

 

DC FieldValueLanguage
dc.contributor.authorLuo, Ping-
dc.contributor.authorWang, Xiaogang-
dc.contributor.authorTang, Xiaoou-
dc.date.accessioned2019-08-12T09:56:18Z-
dc.date.available2019-08-12T09:56:18Z-
dc.date.issued2013-
dc.identifier.citationProceedings of the IEEE International Conference on Computer Vision, 2013, p. 2648-2655-
dc.identifier.urihttp://hdl.handle.net/10722/273661-
dc.description.abstractWe propose a new Deep Decompositional Network (DDN) for parsing pedestrian images into semantic regions, such as hair, head, body, arms, and legs, where the pedestrians can be heavily occluded. Unlike existing methods based on template matching or Bayesian inference, our approach directly maps low-level visual features to the label maps of body parts with DDN, which is able to accurately estimate complex pose variations with good robustness to occlusions and background clutters. DDN jointly estimates occluded regions and segments body parts by stacking three types of hidden layers: occlusion estimation layers, completion layers, and decomposition layers. The occlusion estimation layers estimate a binary mask, indicating which part of a pedestrian is invisible. The completion layers synthesize low-level features of the invisible part from the original features and the occlusion mask. The decomposition layers directly transform the synthesized visual features to label maps. We devise a new strategy to pre-train these hidden layers, and then fine-tune the entire network using the stochastic gradient descent. Experimental results show that our approach achieves better segmentation accuracy than the state-of-the-art methods on pedestrian images with or without occlusions. Another important contribution of this paper is that it provides a large scale benchmark human parsing dataset that includes 3,673 annotated samples collected from 171 surveillance videos. It is 20 times larger than existing public datasets. © 2013 IEEE.-
dc.languageeng-
dc.relation.ispartofProceedings of the IEEE International Conference on Computer Vision-
dc.subjectpedestrian parsing-
dc.subjectdeep learning-
dc.titlePedestrian parsing via deep decompositional network-
dc.typeConference_Paper-
dc.description.naturelink_to_subscribed_fulltext-
dc.identifier.doi10.1109/ICCV.2013.329-
dc.identifier.scopuseid_2-s2.0-84898770979-
dc.identifier.spage2648-
dc.identifier.epage2655-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats