{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,6,18]],"date-time":"2025-06-18T04:10:52Z","timestamp":1750219852978,"version":"3.41.0"},"reference-count":76,"publisher":"Association for Computing Machinery (ACM)","issue":"3","license":[{"start":{"date-parts":[[2023,11,10]],"date-time":"2023-11-10T00:00:00Z","timestamp":1699574400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/www.acm.org\/publications\/policies\/copyright_policy#Background"}],"funder":[{"name":"National Science Foundation","award":["2106965"],"award-info":[{"award-number":["2106965"]}]}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["ACM Trans. Multimedia Comput. Commun. Appl."],"published-print":{"date-parts":[[2024,3,31]]},"abstract":"<jats:p>Extracting roads in aerial images has numerous applications in artificial intelligence and multimedia computing, including traffic pattern analysis and parking space planning. Learning deep neural networks, though very successful, demand vast amounts of high-quality annotations, of which acquisition is time-consuming and expensive. In this work, we propose a semi-supervised approach for image-based road extraction in which only a small set of labeled images are available for training to address this challenge. We design a pixel-wise contrastive loss to self-supervise the network training to utilize the large corpus of unlabeled images. The key idea is to identify pairs of overlapping image regions (positive) or non-overlapping image regions (negative) and encourage the network to make similar outputs for positive pairs or dissimilar outputs for negative pairs. We also develop a negative sampling strategy to filter false-negative samples during the process. An iterative procedure is introduced to apply the network over raw images to generate pseudo-labels, filter and select high-quality labels with the proposed contrastive loss, and retrain the network with the enlarged training dataset. We repeat these iterative steps until convergence. We validate the effectiveness of the proposed methods by performing extensive experiments on the public SpaceNet3 and DeepGlobe Road datasets. Results show that our proposed method achieves state-of-the-art results on public image segmentation benchmarks and significantly outperforms other semi-supervised methods.<\/jats:p>","DOI":"10.1145\/3606374","type":"journal-article","created":{"date-parts":[[2023,7,22]],"date-time":"2023-07-22T09:18:18Z","timestamp":1690017498000},"page":"1-21","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":2,"title":["An Iterative Semi-supervised Approach with Pixel-wise Contrastive Loss for Road Extraction in Aerial Images"],"prefix":"10.1145","volume":"20","author":[{"ORCID":"https:\/\/orcid.org\/0009-0003-1611-8468","authenticated-orcid":false,"given":"Huijie","family":"Zhang","sequence":"first","affiliation":[{"name":"San Diego State University, USA and University of California, Santa Barbara, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0516-2880","authenticated-orcid":false,"given":"Pu","family":"Li","sequence":"additional","affiliation":[{"name":"San Diego State University, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7944-1292","authenticated-orcid":false,"given":"Xiaobai","family":"Liu","sequence":"additional","affiliation":[{"name":"San Diego State University, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9416-6882","authenticated-orcid":false,"given":"Xianfeng","family":"Yang","sequence":"additional","affiliation":[{"name":"Civil and Environmental Engineering, University of Maryland, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7933-5174","authenticated-orcid":false,"given":"Li","family":"An","sequence":"additional","affiliation":[{"name":"San Diego State University, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"320","published-online":{"date-parts":[[2023,11,10]]},"reference":[{"key":"e_1_3_1_2_2","doi-asserted-by":"publisher","DOI":"10.3390\/rs12091444"},{"key":"e_1_3_1_3_2","doi-asserted-by":"publisher","DOI":"10.1145\/3462635"},{"key":"e_1_3_1_4_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00811"},{"key":"e_1_3_1_5_2","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2010.161"},{"key":"e_1_3_1_6_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00496"},{"key":"e_1_3_1_7_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.01063"},{"key":"e_1_3_1_8_2","doi-asserted-by":"publisher","DOI":"10.1145\/1553374.1553380"},{"key":"e_1_3_1_9_2","article-title":"Curriculum labeling: Revisiting pseudo-labeling for semi-supervised learning","author":"Cascante-Bonilla Paola","year":"2020","unstructured":"Paola Cascante-Bonilla, Fuwen Tan, Yanjun Qi, and Vicente Ordonez. 2020. Curriculum labeling: Revisiting pseudo-labeling for semi-supervised learning. arXiv preprint arXiv:2001.06001 (2020).","journal-title":"arXiv preprint arXiv:2001.06001"},{"key":"e_1_3_1_10_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.media.2023.102792"},{"key":"e_1_3_1_11_2","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-58545-7_40"},{"key":"e_1_3_1_12_2","first-page":"1597","volume-title":"International Conference on Machine Learning","author":"Chen Ting","year":"2020","unstructured":"Ting Chen, Simon Kornblith, Mohammad Norouzi, and Geoffrey Hinton. 2020. A simple framework for contrastive learning of visual representations. In International Conference on Machine Learning. PMLR, 1597\u20131607."},{"key":"e_1_3_1_13_2","article-title":"Improved baselines with momentum contrastive learning","author":"Chen Xinlei","year":"2020","unstructured":"Xinlei Chen, Haoqi Fan, Ross Girshick, and Kaiming He. 2020. Improved baselines with momentum contrastive learning. arXiv preprint arXiv:2003.04297 (2020).","journal-title":"arXiv preprint arXiv:2003.04297"},{"key":"e_1_3_1_14_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.00264"},{"key":"e_1_3_1_15_2","doi-asserted-by":"publisher","DOI":"10.1109\/TGRS.2017.2669341"},{"key":"e_1_3_1_16_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2015.191"},{"key":"e_1_3_1_17_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2005.177"},{"key":"e_1_3_1_18_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPRW.2018.00031"},{"key":"e_1_3_1_19_2","article-title":"Improved regularization of convolutional neural networks with Cutout","author":"DeVries Terrance","year":"2017","unstructured":"Terrance DeVries and Graham W. Taylor. 2017. Improved regularization of convolutional neural networks with Cutout. arXiv preprint arXiv:1708.04552 (2017).","journal-title":"arXiv preprint arXiv:1708.04552"},{"issue":"2","key":"e_1_3_1_20_2","first-page":"5","article-title":"Semi-supervised semantic segmentation via dynamic self-training and class-balanced curriculum","volume":"1","author":"Feng Zhengyang","year":"2020","unstructured":"Zhengyang Feng, Qianyu Zhou, Guangliang Cheng, Xin Tan, Jianping Shi, and Lizhuang Ma. 2020. Semi-supervised semantic segmentation via dynamic self-training and class-balanced curriculum. arXiv preprint arXiv:2004.08514 1, 2 (2020), 5.","journal-title":"arXiv preprint arXiv:2004.08514"},{"issue":"4","key":"e_1_3_1_21_2","first-page":"5","article-title":"Consistency regularization and CutMix for semi-supervised semantic segmentation","volume":"2","author":"French Geoffrey","year":"2019","unstructured":"Geoffrey French, Timo Aila, Samuli Laine, Michal Mackiewicz, and Graham Finlayson. 2019. Consistency regularization and CutMix for semi-supervised semantic segmentation. arXiv preprint arXiv:1906.01916 2, 4 (2019), 5.","journal-title":"arXiv preprint arXiv:1906.01916"},{"key":"e_1_3_1_22_2","article-title":"Semi-supervised semantic segmentation needs strong, varied perturbations","author":"French Geoff","year":"2020","unstructured":"Geoff French, S. Laine, Timo Aila, Michal Mackiewicz, and G. Finlayson. 2020. Semi-supervised semantic segmentation needs strong, varied perturbations. arXiv: Computer Vision and Pattern Recognition (2020).","journal-title":"arXiv: Computer Vision and Pattern Recognition"},{"key":"e_1_3_1_23_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.00975"},{"key":"e_1_3_1_24_2","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-58586-0_4"},{"key":"e_1_3_1_25_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2019.00300"},{"issue":"1","key":"e_1_3_1_26_2","first-page":"550","article-title":"AlignSeg: Feature-aligned segmentation networks","volume":"44","author":"Huang Zilong","year":"2021","unstructured":"Zilong Huang, Yunchao Wei, Xinggang Wang, Wenyu Liu, Thomas S. Huang, and Humphrey Shi. 2021. AlignSeg: Feature-aligned segmentation networks. IEEE Transactions on Pattern Analysis and Machine Intelligence 44, 1 (2021), 550\u2013557.","journal-title":"IEEE Transactions on Pattern Analysis and Machine Intelligence"},{"key":"e_1_3_1_27_2","article-title":"Adversarial learning for semi-supervised semantic segmentation","author":"Hung Wei-Chih","year":"2018","unstructured":"Wei-Chih Hung, Yi-Hsuan Tsai, Yan-Ting Liou, Yen-Yu Lin, and Ming-Hsuan Yang. 2018. Adversarial learning for semi-supervised semantic segmentation. arXiv preprint arXiv:1802.07934 (2018).","journal-title":"arXiv preprint arXiv:1802.07934"},{"key":"e_1_3_1_28_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.01273"},{"key":"e_1_3_1_29_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00521"},{"key":"e_1_3_1_30_2","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v33i01.33013967"},{"key":"e_1_3_1_31_2","first-page":"611","article-title":"Remote sensing of urban\/suburban infrastructure and socio-economic attributes","volume":"65","author":"Jensen John R.","year":"1999","unstructured":"John R. Jensen and Dave C. Cowen. 1999. Remote sensing of urban\/suburban infrastructure and socio-economic attributes. Photogrammetric Engineering and Remote Sensing 65 (1999), 611\u2013622.","journal-title":"Photogrammetric Engineering and Remote Sensing"},{"key":"e_1_3_1_32_2","article-title":"Consistency-based semi-supervised learning for object detection","volume":"32","author":"Jeong Jisoo","year":"2019","unstructured":"Jisoo Jeong, Seungeui Lee, Jeesoo Kim, and Nojun Kwak. 2019. Consistency-based semi-supervised learning for object detection. Advances in Neural Information Processing Systems 32 (2019).","journal-title":"Advances in Neural Information Processing Systems"},{"key":"e_1_3_1_33_2","doi-asserted-by":"publisher","DOI":"10.1109\/TGRS.2020.3007029"},{"key":"e_1_3_1_34_2","article-title":"Structured consistency loss for semi-supervised semantic segmentation","author":"Kim Jongmok","year":"2020","unstructured":"Jongmok Kim, Jooyoung Jang, and Hyunwoo Park. 2020. Structured consistency loss for semi-supervised semantic segmentation. arXiv preprint arXiv:2001.04647 (2020).","journal-title":"arXiv preprint arXiv:2001.04647"},{"key":"e_1_3_1_35_2","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-88013-2_8"},{"key":"e_1_3_1_36_2","doi-asserted-by":"publisher","DOI":"10.1109\/TIP.2019.2952690"},{"key":"e_1_3_1_37_2","doi-asserted-by":"publisher","DOI":"10.1145\/3474085.3475602"},{"key":"e_1_3_1_38_2","article-title":"Learning to self-train for semi-supervised few-shot classification","volume":"32","author":"Li Xinzhe","year":"2019","unstructured":"Xinzhe Li, Qianru Sun, Yaoyao Liu, Qin Zhou, Shibao Zheng, Tat-Seng Chua, and Bernt Schiele. 2019. Learning to self-train for semi-supervised few-shot classification. Advances in Neural Information Processing Systems 32 (2019).","journal-title":"Advances in Neural Information Processing Systems"},{"key":"e_1_3_1_39_2","doi-asserted-by":"publisher","DOI":"10.1109\/LGRS.2019.2931928"},{"key":"e_1_3_1_40_2","doi-asserted-by":"publisher","DOI":"10.1145\/3497747"},{"key":"e_1_3_1_41_2","doi-asserted-by":"publisher","DOI":"10.1109\/TGRS.2019.2926397"},{"key":"e_1_3_1_42_2","doi-asserted-by":"publisher","DOI":"10.1109\/LGRS.2017.2704120"},{"key":"e_1_3_1_43_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.isprsjprs.2017.11.014"},{"key":"e_1_3_1_44_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.372"},{"key":"e_1_3_1_45_2","doi-asserted-by":"publisher","DOI":"10.1109\/JSTARS.2015.2443552"},{"key":"e_1_3_1_46_2","article-title":"Semi-supervised semantic segmentation with high-and low-level consistency","author":"Mittal Sudhanshu","year":"2019","unstructured":"Sudhanshu Mittal, Maxim Tatarchenko, and Thomas Brox. 2019. Semi-supervised semantic segmentation with high-and low-level consistency. IEEE Transactions on Pattern Analysis and Machine Intelligence (2019).","journal-title":"IEEE Transactions on Pattern Analysis and Machine Intelligence"},{"key":"e_1_3_1_47_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00331"},{"key":"e_1_3_1_48_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.01269"},{"key":"e_1_3_1_49_2","doi-asserted-by":"publisher","DOI":"10.1109\/LGRS.2018.2870488"},{"key":"e_1_3_1_50_2","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-24574-4_28"},{"key":"e_1_3_1_51_2","doi-asserted-by":"publisher","DOI":"10.1109\/TIT.1965.1053799"},{"key":"e_1_3_1_52_2","first-page":"4","volume-title":"BMVC","author":"Singh Suriya","year":"2018","unstructured":"Suriya Singh, Anil Batra, Guan Pang, Lorenzo Torresani, Saikat Basu, Manohar Paluri, and C. V. Jawahar. 2018. Self-supervised feature learning for semantic segmentation of overhead imagery. In BMVC, Vol. 1. 4."},{"key":"e_1_3_1_53_2","article-title":"A simple semi-supervised learning framework for object detection","author":"Sohn Kihyuk","year":"2020","unstructured":"Kihyuk Sohn, Zizhao Zhang, Chun-Liang Li, Han Zhang, Chen-Yu Lee, and Tomas Pfister. 2020. A simple semi-supervised learning framework for object detection. arXiv preprint arXiv:2005.04757 (2020).","journal-title":"arXiv preprint arXiv:2005.04757"},{"key":"e_1_3_1_54_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.606"},{"issue":"2","key":"e_1_3_1_55_2","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1145\/3478457","article-title":"SADnet: Semi-supervised single image dehazing method based on an attention mechanism","volume":"18","author":"Sun Ziyi","year":"2022","unstructured":"Ziyi Sun, Yunfeng Zhang, Fangxun Bao, Ping Wang, Xunxiang Yao, and Caiming Zhang. 2022. SADnet: Semi-supervised single image dehazing method based on an attention mechanism. ACM Transactions on Multimedia Computing, Communications, and Applications (TOMM) 18, 2 (2022), 1\u201323.","journal-title":"ACM Transactions on Multimedia Computing, Communications, and Applications (TOMM)"},{"key":"e_1_3_1_56_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.00893"},{"key":"e_1_3_1_57_2","article-title":"SpaceNet: A remote sensing dataset and challenge series","author":"Etten Adam Van","year":"2018","unstructured":"Adam Van Etten, Dave Lindenbaum, and Todd M. Bacastow. 2018. SpaceNet: A remote sensing dataset and challenge series. arXiv preprint arXiv:1807.01232 (2018).","journal-title":"arXiv preprint arXiv:1807.01232"},{"key":"e_1_3_1_58_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.00100"},{"key":"e_1_3_1_59_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00721"},{"key":"e_1_3_1_60_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00759"},{"key":"e_1_3_1_61_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.01041"},{"key":"e_1_3_1_62_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00393"},{"key":"e_1_3_1_63_2","doi-asserted-by":"publisher","DOI":"10.1145\/3419842"},{"key":"e_1_3_1_64_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2015.164"},{"key":"e_1_3_1_65_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00305"},{"key":"e_1_3_1_66_2","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2015.2476813"},{"key":"e_1_3_1_67_2","article-title":"Billion-scale semi-supervised learning for image classification","author":"Yalniz I. Zeki","year":"2019","unstructured":"I. Zeki Yalniz, Herv\u00e9 J\u00e9gou, Kan Chen, Manohar Paluri, and Dhruv Mahajan. 2019. Billion-scale semi-supervised learning for image classification. arXiv preprint arXiv:1905.00546 (2019).","journal-title":"arXiv preprint arXiv:1905.00546"},{"key":"e_1_3_1_68_2","doi-asserted-by":"publisher","DOI":"10.1145\/3089249"},{"key":"e_1_3_1_69_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2019.00612"},{"key":"e_1_3_1_70_2","first-page":"18408","article-title":"FlexMatch: Boosting semi-supervised learning with curriculum pseudo labeling","volume":"34","author":"Zhang Bowen","year":"2021","unstructured":"Bowen Zhang, Yidong Wang, Wenxin Hou, Hao Wu, Jindong Wang, Manabu Okumura, and Takahiro Shinozaki. 2021. FlexMatch: Boosting semi-supervised learning with curriculum pseudo labeling. Advances in Neural Information Processing Systems 34 (2021), 18408\u201318419.","journal-title":"Advances in Neural Information Processing Systems"},{"key":"e_1_3_1_71_2","doi-asserted-by":"publisher","DOI":"10.3390\/rs11091017"},{"key":"e_1_3_1_72_2","doi-asserted-by":"publisher","DOI":"10.1109\/LGRS.2018.2802944"},{"key":"e_1_3_1_73_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.01045"},{"key":"e_1_3_1_74_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00718"},{"key":"e_1_3_1_75_2","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v34i07.7000"},{"key":"e_1_3_1_76_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00695"},{"key":"e_1_3_1_77_2","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-01548-9"}],"container-title":["ACM Transactions on Multimedia Computing, Communications, and Applications"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3606374","content-type":"unspecified","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3606374","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,6,17]],"date-time":"2025-06-17T16:47:09Z","timestamp":1750178829000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3606374"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,11,10]]},"references-count":76,"journal-issue":{"issue":"3","published-print":{"date-parts":[[2024,3,31]]}},"alternative-id":["10.1145\/3606374"],"URL":"https:\/\/doi.org\/10.1145\/3606374","relation":{},"ISSN":["1551-6857","1551-6865"],"issn-type":[{"type":"print","value":"1551-6857"},{"type":"electronic","value":"1551-6865"}],"subject":[],"published":{"date-parts":[[2023,11,10]]},"assertion":[{"value":"2022-09-28","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2023-06-21","order":1,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2023-11-10","order":2,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}