Deformable Convolutional Networks. Dai, J., Qi, H., Xiong, Y., Li, Y., Zhang, G., Hu, H., & Wei, Y. Proceedings of the IEEE International Conference on Computer Vision, 2017-Octob:764-773, 2017. ![Deformable Convolutional Networks [pdf]](https://bibbase.org/img/filetypes/pdf.svg "pdf")
Paper doi abstract bibtex Convolutional neural networks (CNNs) are inherently limited to model geometric transformations due to the fixed geometric structures in their building modules. In this work, we introduce two new modules to enhance the transformation modeling capability of CNNs, namely, deformable convolution and deformable RoI pooling. Both are based on the idea of augmenting the spatial sampling locations in the modules with additional offsets and learning the offsets from the target tasks, without additional supervision. The new modules can readily replace their plain counterparts in existing CNNs and can be easily trained end-to-end by standard back-propagation, giving rise to deformable convolutional networks. Extensive experiments validate the performance of our approach. For the first time, we show that learning dense spatial transformation in deep CNNs is effective for sophisticated vision tasks such as object detection and semantic segmentation. The code is released at https://github.com/msracver/Deformable-ConvNets.
@article{
title = {Deformable Convolutional Networks},
type = {article},
year = {2017},
pages = {764-773},
volume = {2017-Octob},
id = {7245b395-e44e-3f87-9c82-da8191ec86d7},
created = {2021-10-14T06:49:26.004Z},
file_attached = {true},
profile_id = {48fc0258-023d-3602-860e-824092d62c56},
group_id = {1ff583c0-be37-34fa-9c04-73c69437d354},
last_modified = {2021-10-14T06:49:32.317Z},
read = {false},
starred = {false},
authored = {false},
confirmed = {true},
hidden = {false},
folder_uuids = {d54ba66b-a8cf-41de-8e2d-c3256f322e07},
private_publication = {false},
abstract = {Convolutional neural networks (CNNs) are inherently limited to model geometric transformations due to the fixed geometric structures in their building modules. In this work, we introduce two new modules to enhance the transformation modeling capability of CNNs, namely, deformable convolution and deformable RoI pooling. Both are based on the idea of augmenting the spatial sampling locations in the modules with additional offsets and learning the offsets from the target tasks, without additional supervision. The new modules can readily replace their plain counterparts in existing CNNs and can be easily trained end-to-end by standard back-propagation, giving rise to deformable convolutional networks. Extensive experiments validate the performance of our approach. For the first time, we show that learning dense spatial transformation in deep CNNs is effective for sophisticated vision tasks such as object detection and semantic segmentation. The code is released at https://github.com/msracver/Deformable-ConvNets.},
bibtype = {article},
author = {Dai, Jifeng and Qi, Haozhi and Xiong, Yuwen and Li, Yi and Zhang, Guodong and Hu, Han and Wei, Yichen},
doi = {10.1109/ICCV.2017.89},
journal = {Proceedings of the IEEE International Conference on Computer Vision}
}
Downloads: 0
{"_id":"AFDJnQnBJtGLLnmxE","bibbaseid":"dai-qi-xiong-li-zhang-hu-wei-deformableconvolutionalnetworks-2017","authorIDs":[],"author_short":["Dai, J.","Qi, H.","Xiong, Y.","Li, Y.","Zhang, G.","Hu, H.","Wei, Y."],"bibdata":{"title":"Deformable Convolutional Networks","type":"article","year":"2017","pages":"764-773","volume":"2017-Octob","id":"7245b395-e44e-3f87-9c82-da8191ec86d7","created":"2021-10-14T06:49:26.004Z","file_attached":"true","profile_id":"48fc0258-023d-3602-860e-824092d62c56","group_id":"1ff583c0-be37-34fa-9c04-73c69437d354","last_modified":"2021-10-14T06:49:32.317Z","read":false,"starred":false,"authored":false,"confirmed":"true","hidden":false,"folder_uuids":"d54ba66b-a8cf-41de-8e2d-c3256f322e07","private_publication":false,"abstract":"Convolutional neural networks (CNNs) are inherently limited to model geometric transformations due to the fixed geometric structures in their building modules. In this work, we introduce two new modules to enhance the transformation modeling capability of CNNs, namely, deformable convolution and deformable RoI pooling. Both are based on the idea of augmenting the spatial sampling locations in the modules with additional offsets and learning the offsets from the target tasks, without additional supervision. The new modules can readily replace their plain counterparts in existing CNNs and can be easily trained end-to-end by standard back-propagation, giving rise to deformable convolutional networks. Extensive experiments validate the performance of our approach. For the first time, we show that learning dense spatial transformation in deep CNNs is effective for sophisticated vision tasks such as object detection and semantic segmentation. The code is released at https://github.com/msracver/Deformable-ConvNets.","bibtype":"article","author":"Dai, Jifeng and Qi, Haozhi and Xiong, Yuwen and Li, Yi and Zhang, Guodong and Hu, Han and Wei, Yichen","doi":"10.1109/ICCV.2017.89","journal":"Proceedings of the IEEE International Conference on Computer Vision","bibtex":"@article{\n title = {Deformable Convolutional Networks},\n type = {article},\n year = {2017},\n pages = {764-773},\n volume = {2017-Octob},\n id = {7245b395-e44e-3f87-9c82-da8191ec86d7},\n created = {2021-10-14T06:49:26.004Z},\n file_attached = {true},\n profile_id = {48fc0258-023d-3602-860e-824092d62c56},\n group_id = {1ff583c0-be37-34fa-9c04-73c69437d354},\n last_modified = {2021-10-14T06:49:32.317Z},\n read = {false},\n starred = {false},\n authored = {false},\n confirmed = {true},\n hidden = {false},\n folder_uuids = {d54ba66b-a8cf-41de-8e2d-c3256f322e07},\n private_publication = {false},\n abstract = {Convolutional neural networks (CNNs) are inherently limited to model geometric transformations due to the fixed geometric structures in their building modules. In this work, we introduce two new modules to enhance the transformation modeling capability of CNNs, namely, deformable convolution and deformable RoI pooling. Both are based on the idea of augmenting the spatial sampling locations in the modules with additional offsets and learning the offsets from the target tasks, without additional supervision. The new modules can readily replace their plain counterparts in existing CNNs and can be easily trained end-to-end by standard back-propagation, giving rise to deformable convolutional networks. Extensive experiments validate the performance of our approach. For the first time, we show that learning dense spatial transformation in deep CNNs is effective for sophisticated vision tasks such as object detection and semantic segmentation. The code is released at https://github.com/msracver/Deformable-ConvNets.},\n bibtype = {article},\n author = {Dai, Jifeng and Qi, Haozhi and Xiong, Yuwen and Li, Yi and Zhang, Guodong and Hu, Han and Wei, Yichen},\n doi = {10.1109/ICCV.2017.89},\n journal = {Proceedings of the IEEE International Conference on Computer Vision}\n}","author_short":["Dai, J.","Qi, H.","Xiong, Y.","Li, Y.","Zhang, G.","Hu, H.","Wei, Y."],"urls":{"Paper":"https://bibbase.org/service/mendeley/bfbbf840-4c42-3914-a463-19024f50b30c/file/22d26132-b090-db8f-1f61-f38246a7f90e/Deformable_Convolutional_Networks.pdf.pdf"},"biburl":"https://bibbase.org/service/mendeley/bfbbf840-4c42-3914-a463-19024f50b30c","bibbaseid":"dai-qi-xiong-li-zhang-hu-wei-deformableconvolutionalnetworks-2017","role":"author","metadata":{"authorlinks":{}},"downloads":0},"bibtype":"article","biburl":"https://bibbase.org/service/mendeley/bfbbf840-4c42-3914-a463-19024f50b30c","creationDate":"2019-09-15T15:10:41.135Z","downloads":0,"keywords":[],"search_terms":["deformable","convolutional","networks","dai","qi","xiong","li","zhang","hu","wei"],"title":"Deformable Convolutional Networks","year":2017,"dataSources":["rsZGjfhq5jNHuLCRm","ya2CyA73rpZseyrZ8","6yXn8CtuzyEbCSr2m","2252seNhipfTmjEBQ"]}