SegNetRes-CRF: A Deep Convolutional Encoder-Decoder Architecture for Semantic Image Segmentation. De Oliveira Junior, L., Medeiros, H., Macedo, D., Zanchettin, C., Oliveira, A., & Ludermir, T. In Proceedings of the International Joint Conference on Neural Networks, volume 2018-July, 2018. doi abstract bibtex © 2018 IEEE. Semantic segmentation is an essential task in computer vision that aims to label each image pixel. Several of the actual best approaches in this context are based on deep neural networks. For example, SegNet is a deep encoderdecoder architecture approach whose results were disruptive because it is fast and performs well. However, this architecture fails to finedelineating the edges between the objects of interest in the image. We propose some modifications in the SegNet-Basic architecture by using a post-processing segmentation layer (using Conditional Random Fields) and by transferring high resolution features combined to the decoder network. The proposed method was evaluated in the dataset CamVid. Moreover, it was compared with important variants of SegNet and showed to be able to improve the overall accuracy of SegNet-Basic by up to 17.5%.
@inproceedings{
title = {SegNetRes-CRF: A Deep Convolutional Encoder-Decoder Architecture for Semantic Image Segmentation},
type = {inproceedings},
year = {2018},
keywords = {CRF,Encoder-decoder architectures,Image segmentation,Residual networks,SegNet,U-Net},
volume = {2018-July},
id = {8f4e3e07-0c35-3589-a26c-60f718ae8cf7},
created = {2019-02-14T18:02:01.791Z},
file_attached = {false},
profile_id = {74e7d4ea-3dac-3118-aab9-511a5b337e8f},
last_modified = {2019-02-14T18:02:01.791Z},
read = {false},
starred = {false},
authored = {true},
confirmed = {false},
hidden = {false},
private_publication = {false},
abstract = {© 2018 IEEE. Semantic segmentation is an essential task in computer vision that aims to label each image pixel. Several of the actual best approaches in this context are based on deep neural networks. For example, SegNet is a deep encoderdecoder architecture approach whose results were disruptive because it is fast and performs well. However, this architecture fails to finedelineating the edges between the objects of interest in the image. We propose some modifications in the SegNet-Basic architecture by using a post-processing segmentation layer (using Conditional Random Fields) and by transferring high resolution features combined to the decoder network. The proposed method was evaluated in the dataset CamVid. Moreover, it was compared with important variants of SegNet and showed to be able to improve the overall accuracy of SegNet-Basic by up to 17.5%.},
bibtype = {inproceedings},
author = {De Oliveira Junior, L.A. and Medeiros, H.R. and Macedo, D. and Zanchettin, C. and Oliveira, A.L.I. and Ludermir, T.},
doi = {10.1109/IJCNN.2018.8489376},
booktitle = {Proceedings of the International Joint Conference on Neural Networks}
}
Downloads: 0
{"_id":"gwDvznkrbsN4bWxag","bibbaseid":"deoliveirajunior-medeiros-macedo-zanchettin-oliveira-ludermir-segnetrescrfadeepconvolutionalencoderdecoderarchitectureforsemanticimagesegmentation-2018","authorIDs":["95PhW7tkuv95vtHAq","PtDsdiZ3iPSFZKH6J"],"author_short":["De Oliveira Junior, L.","Medeiros, H.","Macedo, D.","Zanchettin, C.","Oliveira, A.","Ludermir, T."],"bibdata":{"title":"SegNetRes-CRF: A Deep Convolutional Encoder-Decoder Architecture for Semantic Image Segmentation","type":"inproceedings","year":"2018","keywords":"CRF,Encoder-decoder architectures,Image segmentation,Residual networks,SegNet,U-Net","volume":"2018-July","id":"8f4e3e07-0c35-3589-a26c-60f718ae8cf7","created":"2019-02-14T18:02:01.791Z","file_attached":false,"profile_id":"74e7d4ea-3dac-3118-aab9-511a5b337e8f","last_modified":"2019-02-14T18:02:01.791Z","read":false,"starred":false,"authored":"true","confirmed":false,"hidden":false,"private_publication":false,"abstract":"© 2018 IEEE. Semantic segmentation is an essential task in computer vision that aims to label each image pixel. Several of the actual best approaches in this context are based on deep neural networks. For example, SegNet is a deep encoderdecoder architecture approach whose results were disruptive because it is fast and performs well. However, this architecture fails to finedelineating the edges between the objects of interest in the image. We propose some modifications in the SegNet-Basic architecture by using a post-processing segmentation layer (using Conditional Random Fields) and by transferring high resolution features combined to the decoder network. The proposed method was evaluated in the dataset CamVid. Moreover, it was compared with important variants of SegNet and showed to be able to improve the overall accuracy of SegNet-Basic by up to 17.5%.","bibtype":"inproceedings","author":"De Oliveira Junior, L.A. and Medeiros, H.R. and Macedo, D. and Zanchettin, C. and Oliveira, A.L.I. and Ludermir, T.","doi":"10.1109/IJCNN.2018.8489376","booktitle":"Proceedings of the International Joint Conference on Neural Networks","bibtex":"@inproceedings{\n title = {SegNetRes-CRF: A Deep Convolutional Encoder-Decoder Architecture for Semantic Image Segmentation},\n type = {inproceedings},\n year = {2018},\n keywords = {CRF,Encoder-decoder architectures,Image segmentation,Residual networks,SegNet,U-Net},\n volume = {2018-July},\n id = {8f4e3e07-0c35-3589-a26c-60f718ae8cf7},\n created = {2019-02-14T18:02:01.791Z},\n file_attached = {false},\n profile_id = {74e7d4ea-3dac-3118-aab9-511a5b337e8f},\n last_modified = {2019-02-14T18:02:01.791Z},\n read = {false},\n starred = {false},\n authored = {true},\n confirmed = {false},\n hidden = {false},\n private_publication = {false},\n abstract = {© 2018 IEEE. Semantic segmentation is an essential task in computer vision that aims to label each image pixel. Several of the actual best approaches in this context are based on deep neural networks. For example, SegNet is a deep encoderdecoder architecture approach whose results were disruptive because it is fast and performs well. However, this architecture fails to finedelineating the edges between the objects of interest in the image. We propose some modifications in the SegNet-Basic architecture by using a post-processing segmentation layer (using Conditional Random Fields) and by transferring high resolution features combined to the decoder network. The proposed method was evaluated in the dataset CamVid. Moreover, it was compared with important variants of SegNet and showed to be able to improve the overall accuracy of SegNet-Basic by up to 17.5%.},\n bibtype = {inproceedings},\n author = {De Oliveira Junior, L.A. and Medeiros, H.R. and Macedo, D. and Zanchettin, C. and Oliveira, A.L.I. and Ludermir, T.},\n doi = {10.1109/IJCNN.2018.8489376},\n booktitle = {Proceedings of the International Joint Conference on Neural Networks}\n}","author_short":["De Oliveira Junior, L.","Medeiros, H.","Macedo, D.","Zanchettin, C.","Oliveira, A.","Ludermir, T."],"biburl":"https://bibbase.org/service/mendeley/74e7d4ea-3dac-3118-aab9-511a5b337e8f","bibbaseid":"deoliveirajunior-medeiros-macedo-zanchettin-oliveira-ludermir-segnetrescrfadeepconvolutionalencoderdecoderarchitectureforsemanticimagesegmentation-2018","role":"author","urls":{},"keyword":["CRF","Encoder-decoder architectures","Image segmentation","Residual networks","SegNet","U-Net"],"metadata":{"authorlinks":{"zanchettin, c":"https://bibbase.org/service/mendeley/74e7d4ea-3dac-3118-aab9-511a5b337e8f","zanchettin, c":"https://zanche.github.io/publications/"}},"downloads":0},"bibtype":"inproceedings","creationDate":"2020-09-17T14:33:46.314Z","downloads":0,"keywords":["crf","encoder-decoder architectures","image segmentation","residual networks","segnet","u-net"],"search_terms":["segnetres","crf","deep","convolutional","encoder","decoder","architecture","semantic","image","segmentation","de oliveira junior","medeiros","macedo","zanchettin","oliveira","ludermir"],"title":"SegNetRes-CRF: A Deep Convolutional Encoder-Decoder Architecture for Semantic Image Segmentation","year":2018,"biburl":"https://bibbase.org/service/mendeley/74e7d4ea-3dac-3118-aab9-511a5b337e8f","dataSources":["fvRdkx56Jpp5ebtSw","XkGKCoQgZDKqXZqdh","ya2CyA73rpZseyrZ8","2252seNhipfTmjEBQ"]}