File Download
Supplementary

Conference Paper: Visual saliency based on multiscale deep features

TitleVisual saliency based on multiscale deep features
Authors
KeywordsComputer vision
Pattern recognition
Issue Date2015
Citation
The 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Boston, MA., 7-12 June 2015. In Conference Proceedings, 2015, p. 1-9 How to Cite?
AbstractVisual saliency is a fundamental problem in both cognitive and computational sciences, including computer vision. In this CVPR 2015 paper, we discover that a high-quality visual saliency model can be trained with multiscale features extracted using a popular deep learning architecture, convolutional neural networks (CNNs), which have had many successes in visual recognition tasks. For learning such saliency models, we introduce a neural network architecture, which has fully connected layers on top of CNNs responsible for extracting features at three different scales. We then propose a refinement method to enhance the spatial coherence of our saliency results. Finally, aggregating multiple saliency maps computed for different levels of image segmentation can further boost the performance, yielding saliency maps better than those generated from a single segmentation. To promote further research and evaluation of visual saliency models, we also construct a new large database of 4447 challenging images and their pixelwise saliency annotation. Experimental results demonstrate that our proposed method is capable of achieving state-of-the-art performance on all public benchmarks, improving the F-Measure by 5.0% and 13.2% respectively on the MSRA-B dataset and our new dataset (HKU-IS), and lowering the mean absolute error by 5.7% and 35.1% respectively on these two datasets.
Persistent Identifierhttp://hdl.handle.net/10722/214077

 

DC FieldValueLanguage
dc.contributor.authorLi, G-
dc.contributor.authorYu, Y-
dc.date.accessioned2015-08-20T01:59:33Z-
dc.date.available2015-08-20T01:59:33Z-
dc.date.issued2015-
dc.identifier.citationThe 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Boston, MA., 7-12 June 2015. In Conference Proceedings, 2015, p. 1-9-
dc.identifier.urihttp://hdl.handle.net/10722/214077-
dc.description.abstractVisual saliency is a fundamental problem in both cognitive and computational sciences, including computer vision. In this CVPR 2015 paper, we discover that a high-quality visual saliency model can be trained with multiscale features extracted using a popular deep learning architecture, convolutional neural networks (CNNs), which have had many successes in visual recognition tasks. For learning such saliency models, we introduce a neural network architecture, which has fully connected layers on top of CNNs responsible for extracting features at three different scales. We then propose a refinement method to enhance the spatial coherence of our saliency results. Finally, aggregating multiple saliency maps computed for different levels of image segmentation can further boost the performance, yielding saliency maps better than those generated from a single segmentation. To promote further research and evaluation of visual saliency models, we also construct a new large database of 4447 challenging images and their pixelwise saliency annotation. Experimental results demonstrate that our proposed method is capable of achieving state-of-the-art performance on all public benchmarks, improving the F-Measure by 5.0% and 13.2% respectively on the MSRA-B dataset and our new dataset (HKU-IS), and lowering the mean absolute error by 5.7% and 35.1% respectively on these two datasets.-
dc.languageeng-
dc.relation.ispartofConference on Computer Vision and Pattern Recognition-
dc.rightsCreative Commons: Attribution 3.0 Hong Kong License-
dc.subjectComputer vision-
dc.subjectPattern recognition-
dc.titleVisual saliency based on multiscale deep features-
dc.typeConference_Paper-
dc.identifier.emailYu, Y: yzyu@cs.hku.hk-
dc.identifier.authorityYu, Y=rp01415-
dc.description.naturepublished_or_final_version-
dc.identifier.hkuros249497-
dc.identifier.spage1-
dc.identifier.epage9-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats