{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,30]],"date-time":"2026-04-30T14:40:16Z","timestamp":1777560016347,"version":"3.51.4"},"reference-count":46,"publisher":"SAGE Publications","issue":"1","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["AIC"],"published-print":{"date-parts":[[2023,2,9]]},"abstract":"<jats:p>Recent years have witnessed a rapid advance in training and testing synthetic data through deep learning networks for the annotation of synthetic data that can be automatically marked. However, a domain discrepancy still exists between synthetic data and real data. In this paper, we address the domain discrepancy issue from three aspects: 1) We design a synthetic image generator with automatically labeled based on 3D scenes. 2) A novel adversarial domain adaptation model is proposed to learn robust intermediate representation free of distractors to improve the transfer performance. 3) We construct a distractor-invariant network and adopt the sample transferability strategy on global-local levels, respectively, to mitigate the cross-domain gap. Additional exploratory experiments demonstrate that the proposed model achieves large performance margins, which show significant advance over the other state-of-the-art models, performing a promotion of 10%\u201315% mAP on various domain adaptation scenarios.<\/jats:p>","DOI":"10.3233\/aic-220039","type":"journal-article","created":{"date-parts":[[2022,12,30]],"date-time":"2022-12-30T11:39:07Z","timestamp":1672400347000},"page":"13-25","source":"Crossref","is-referenced-by-count":2,"title":["Learning invariant representation using synthetic imagery for object detection"],"prefix":"10.1177","volume":"36","author":[{"given":"Ning","family":"Jiang","sequence":"first","affiliation":[{"name":"School of Computer Science and Technology, Hangzhou Dianzi University, Hangzhou, China"},{"name":"School of Information and Intelligent Engineering, Ningbo City College of Vocational Technology, Ningbo, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jinglong","family":"Fang","sequence":"additional","affiliation":[{"name":"School of Computer Science and Technology, Hangzhou Dianzi University, Hangzhou, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yanli","family":"Shao","sequence":"additional","affiliation":[{"name":"School of Computer Science and Technology, Hangzhou Dianzi University, Hangzhou, China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"179","reference":[{"key":"10.3233\/AIC-220039_ref1","doi-asserted-by":"crossref","unstructured":"P.\u00a0Bateni, R.\u00a0Goyal, U.\u00a0Franke and S.R.\u00a0Improved, Few-shot visual classification, in: CVPR, 2020.","DOI":"10.1109\/CVPR42600.2020.01450"},{"key":"10.3233\/AIC-220039_ref2","doi-asserted-by":"crossref","unstructured":"H.\u00a0Bilen and A.\u00a0Vedaldi, Weakly supervised deep detection networks, in: CVPR, 2016.","DOI":"10.1109\/CVPR.2016.311"},{"key":"10.3233\/AIC-220039_ref3","doi-asserted-by":"crossref","unstructured":"P.P.\u00a0Busto and J.\u00a0Gall, Open set domain adaptation, in: ICCV, 2017.","DOI":"10.1109\/ICCV.2017.88"},{"key":"10.3233\/AIC-220039_ref4","doi-asserted-by":"crossref","unstructured":"Q.\u00a0Cai, Y.\u00a0Pan, C.-W.\u00a0Ngo, X.\u00a0Tian, L.\u00a0Duan and T.\u00a0Yao, Exploring object relation in mean teacher for cross-domain detection, in: CVPR, 2019.","DOI":"10.1109\/CVPR.2019.01172"},{"key":"10.3233\/AIC-220039_ref5","doi-asserted-by":"crossref","unstructured":"Y.\u00a0Cao, K.\u00a0Chen, C.\u00a0Change Loy and D.\u00a0Lin, Prime sample attention in object detection, in: Computer Vision and Pattern Recognition (CVPR), 2020, pp.\u00a01435\u20131447.","DOI":"10.1109\/CVPR42600.2020.01160"},{"key":"10.3233\/AIC-220039_ref6","doi-asserted-by":"crossref","unstructured":"Y.\u00a0Chen, W.\u00a0Li, C.\u00a0Sakaridis, D.\u00a0Dai and L.\u00a0Van Gool, Domain adaptive faster r-cnn for object detection in the wild, in: CVPR, 2018.","DOI":"10.1109\/CVPR.2018.00352"},{"key":"10.3233\/AIC-220039_ref7","doi-asserted-by":"crossref","unstructured":"M.\u00a0Cordts, M.\u00a0Omran, S.\u00a0Ramos, T.\u00a0Redfield, M.\u00a0Enzweiler, R.\u00a0Berenson, U.\u00a0Franke, S.\u00a0Roth and B.\u00a0Schiele, The cityscapes dataset for semantic urban scene understanding, in: CVPR, 2016.","DOI":"10.1109\/CVPR.2016.350"},{"key":"10.3233\/AIC-220039_ref8","unstructured":"C.\u00a0Cortes, Y.\u00a0Mansour and M.\u00a0Mohri, Learning bounds for importance weighting, in: Neur IPS, 2010."},{"key":"10.3233\/AIC-220039_ref10","doi-asserted-by":"crossref","unstructured":"G.\u00a0Csurka, B.\u00a0Chidlovskii and S.\u00a0Clinchant, Adapted domain specific class means, in: Proceedings of the IEEE International Conference on Computer Vision Workshops, IEEE, Santiago, Chile, 2015, pp.\u00a07\u201311.","DOI":"10.1109\/ICCVW.2015.20"},{"issue":"2","key":"10.3233\/AIC-220039_ref13","doi-asserted-by":"publisher","first-page":"303","DOI":"10.1007\/s11263-009-0275-4","article-title":"The Pascal visual object classes (voc) challenge","volume":"88","author":"Everingham","year":"2010","journal-title":"IJCV"},{"issue":"59","key":"10.3233\/AIC-220039_ref14","first-page":"1","article-title":"Domain adversarial training of neural networks","volume":"17","author":"Ganin","year":"2016","journal-title":"JMLR"},{"key":"10.3233\/AIC-220039_ref15","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2015.169"},{"key":"10.3233\/AIC-220039_ref17","unstructured":"I.\u00a0Goodfellow, J.\u00a0Pouget, B.\u00a0Xu et al., Generative adversarial nets, in: Proceedings of the 2014 IEEE Conference on Computer Vision and Pattern Recognition, IEEE, Piscataway, 2014, pp.\u00a0367\u2013384."},{"key":"10.3233\/AIC-220039_ref18","unstructured":"J.\u00a0Hoffman, E.\u00a0Tzeng, T.\u00a0Park, J.-Y.\u00a0Zhu, P.\u00a0Isola, K.\u00a0Saenko, A.A.\u00a0Efros and T.\u00a0Darrell, Cycada: Cycle-consistent adversarial domain adaptation, in: ICML 2018, 2018."},{"key":"10.3233\/AIC-220039_ref19","doi-asserted-by":"crossref","unstructured":"H.-K.\u00a0Hsu, W.-C.\u00a0Hung, H.-Y.\u00a0Tseng, C.-H.\u00a0Yao, Y.-H.\u00a0Tsai, M.\u00a0Singh and M.-H.\u00a0Yang, Progressive domain adaptation for object detection, in: CVPR Workshops, 2019.","DOI":"10.1109\/WACV45572.2020.9093358"},{"key":"10.3233\/AIC-220039_ref21","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.632"},{"key":"10.3233\/AIC-220039_ref22","doi-asserted-by":"crossref","unstructured":"Y.\u00a0Jia, E.\u00a0Shelhamer, J.\u00a0Donahue, S.\u00a0Karayev, J.\u00a0Long, R.\u00a0Girshick, S.\u00a0Guadarrama and T.\u00a0Darrell, Caffe: Convolutional architecture for fast feature embedding, in: ACMMM, 2014.","DOI":"10.1145\/2647868.2654889"},{"key":"10.3233\/AIC-220039_ref24","doi-asserted-by":"crossref","unstructured":"M.\u00a0Khodabandeh, A.\u00a0Vahdat, M.\u00a0Ranjbar and W.G.\u00a0Macready, A robust learning approach to domain adaptive object detection, in: ICCV, 2019.","DOI":"10.1109\/ICCV.2019.00057"},{"key":"10.3233\/AIC-220039_ref25","doi-asserted-by":"crossref","unstructured":"M.\u00a0Kim and H.\u00a0Byun, Learning texture invariant representation for domain adaptation of semantic segmentation, in: Computer Vision and Pattern Recognition (CVPR), 2020, pp.\u00a01324\u20131335.","DOI":"10.1109\/CVPR42600.2020.01299"},{"key":"10.3233\/AIC-220039_ref26","doi-asserted-by":"crossref","unstructured":"S.\u00a0Kim, J.\u00a0Choi, T.\u00a0Kim and C.\u00a0Kim, Self-training and adversarial background regularization for unsupervised domain adaptive one-stage object detection, in: ICCV, 2019.","DOI":"10.1109\/ICCV.2019.00619"},{"key":"10.3233\/AIC-220039_ref27","doi-asserted-by":"crossref","unstructured":"T.\u00a0Kim, M.\u00a0Jeong, S.\u00a0Kim, S.\u00a0Choi and C.\u00a0Kim, Diversify and match: A domain adaptive representation learning paradigm for object detection, in: CVPR, 2019.","DOI":"10.1109\/CVPR.2019.01274"},{"key":"10.3233\/AIC-220039_ref28","unstructured":"A.\u00a0Krizhevsky, I.\u00a0Sutskever and G.E.\u00a0Hinton, Imagenet classification with deep convolutional neural networks, in: NIPS, 2012."},{"key":"10.3233\/AIC-220039_ref29","doi-asserted-by":"crossref","unstructured":"T.-Y.\u00a0Lin, P.\u00a0Goyal, R.\u00a0Girshick, K.\u00a0He and P.\u00a0Dollar, Focal loss for dense object detection, in: ICCV, 2017.","DOI":"10.1109\/ICCV.2017.324"},{"key":"10.3233\/AIC-220039_ref30","unstructured":"T.-Y.\u00a0Lin, M.\u00a0Maire, S.\u00a0Belongie, J.\u00a0Hays, P.\u00a0Perona, D.\u00a0Ramanan, P.\u00a0Doll\u00e1r and C.L.\u00a0Zitnick, Visual commonsense R-CNN, in: ECCV, 2020."},{"key":"10.3233\/AIC-220039_ref31","unstructured":"M.-Y.\u00a0Liu, T.\u00a0Breuel and J.\u00a0Kautz, Unsupervised image-to image translation networks, in: NIPS, 2017."},{"key":"10.3233\/AIC-220039_ref32","doi-asserted-by":"crossref","unstructured":"W.\u00a0Liu, D.\u00a0Anguelov, D.\u00a0Erhan, C.\u00a0Szegedy, S.\u00a0Reed, C.-Y.\u00a0Fu and A.C.\u00a0Berg, Ssd: Single shot multibox detector, in: ECCV, 2016.","DOI":"10.1007\/978-3-319-46448-0_2"},{"key":"10.3233\/AIC-220039_ref33","unstructured":"M.\u00a0Long, H.\u00a0Zhu, J.\u00a0Wang and M.I.\u00a0Jordan, Unsupervised domain adaptation with residual transfer networks, in: NIPS, 2016."},{"key":"10.3233\/AIC-220039_ref34","doi-asserted-by":"crossref","unstructured":"F.\u00a0Maria Carlucci, L.\u00a0Porzi, B.\u00a0Caputo, E.\u00a0Ricci and S.\u00a0Rota Bul\u2018o, Autodial: Automatic domain alignment layers, in: International Conference on Computer Vision, 2017.","DOI":"10.1109\/ICCV.2017.542"},{"key":"10.3233\/AIC-220039_ref35","unstructured":"A.\u00a0Odena, C.\u00a0Olah and J.\u00a0Shlens, Unbiased Scene Graph Generation from Biased Training, 2020."},{"key":"10.3233\/AIC-220039_ref37","doi-asserted-by":"crossref","unstructured":"J.\u00a0Redmon, S.\u00a0Divvala, R.\u00a0Girshick and A.\u00a0Farhadi, You only look once: Unified, real-time object detection, in: CVPR, 2016.","DOI":"10.1109\/CVPR.2016.91"},{"key":"10.3233\/AIC-220039_ref38","unstructured":"S.\u00a0Ren, K.\u00a0He, R.\u00a0Girshick and J.\u00a0Sun, Faster r-cnn: Towards real-time object detection with region proposal networks, in: NIPS, 2015."},{"key":"10.3233\/AIC-220039_ref39","doi-asserted-by":"crossref","unstructured":"S.R.\u00a0Richter, V.\u00a0Vineet, S.\u00a0Roth and V.\u00a0Koltun, Playing for data: Ground truth from computer games, in: European Conference on Computer Vision, Springer, 2016, pp.\u00a0102\u2013118.","DOI":"10.1007\/978-3-319-46475-6_7"},{"key":"10.3233\/AIC-220039_ref40","unstructured":"K.\u00a0Saenko, B.\u00a0Kulis, M.\u00a0Fritz and T.\u00a0Darrell, Meta-transfer learning for zero-shot super-resolution, in: ECCV, 2020."},{"key":"10.3233\/AIC-220039_ref41","unstructured":"K.\u00a0Saito, Y.\u00a0Ushiku, T.\u00a0Harada and K.\u00a0Saenko, Adversarial dropout regularization, in: ICLR, 2018."},{"key":"10.3233\/AIC-220039_ref42","doi-asserted-by":"crossref","unstructured":"K.\u00a0Saito, Y.\u00a0Ushiku, T.\u00a0Harada and K.\u00a0Saenko, Strongweak distribution alignment for adaptive object detection, in: CVPR, 2019.","DOI":"10.1109\/CVPR.2019.00712"},{"key":"10.3233\/AIC-220039_ref43","doi-asserted-by":"crossref","unstructured":"K.\u00a0Saito, K.\u00a0Watanabe, Y.\u00a0Ushiku and T.\u00a0Harada, Maximum classifier discrepancy for unsupervised domain adaptation, in: CVPR, 2018.","DOI":"10.1109\/CVPR.2018.00392"},{"key":"10.3233\/AIC-220039_ref44","doi-asserted-by":"crossref","unstructured":"K.\u00a0Saito, S.\u00a0Yamamoto, Y.\u00a0Ushiku and T.\u00a0Harada, Open set domain adaptation by backpropagation, in: ICCV, 2018.","DOI":"10.1007\/978-3-030-01228-1_10"},{"key":"10.3233\/AIC-220039_ref45","doi-asserted-by":"crossref","unstructured":"C.\u00a0Sakaridis, D.\u00a0Dai and L.\u00a0Van Gool, Semantic foggy scene understanding with synthetic data, in: IJCV, 2018.","DOI":"10.1007\/s11263-018-1072-8"},{"key":"10.3233\/AIC-220039_ref46","doi-asserted-by":"crossref","unstructured":"S.\u00a0Sankaranarayanan, Y.\u00a0Balaji, A.\u00a0Jain, S.N.\u00a0Lim and R.\u00a0Chellappa, Learning from synthetic data: Addressing domain shift for semantic segmentation, in: CVPR, 2018.","DOI":"10.1109\/CVPR.2018.00395"},{"key":"10.3233\/AIC-220039_ref47","doi-asserted-by":"crossref","unstructured":"Y.\u00a0Tang, J.\u00a0Wang, B.\u00a0Gao, E.\u00a0Dellandr\u00e9a, R.\u00a0Gaizauskas and L.\u00a0Chen, Large scale semi-supervised object detection using visual and semantic knowledge transfer, in: CVPR, 2016.","DOI":"10.1109\/CVPR.2016.233"},{"key":"10.3233\/AIC-220039_ref48","doi-asserted-by":"crossref","unstructured":"K.\u00a0Tian, C.\u00a0Zhang, Y.\u00a0Wang, S.\u00a0Xiang and C.\u00a0Pan, Knowledge mining and transferring for domain adaptive object detection, in: Proceedings of the IEEE\/CVF International Conference on Computer Vision (ICCV), 2021, pp.\u00a09133\u20139142.","DOI":"10.1109\/ICCV48922.2021.00900"},{"key":"10.3233\/AIC-220039_ref49","doi-asserted-by":"crossref","unstructured":"E.\u00a0Tzeng, J.\u00a0Hoffman, K.\u00a0Saenko and T.\u00a0Darrell, Adversarial discriminative domain adaptation, in: CVPR, 2017.","DOI":"10.1109\/CVPR.2017.316"},{"key":"10.3233\/AIC-220039_ref50","unstructured":"V.S.\u00a0Vibashan, V.\u00a0Gupta, P.\u00a0Oza, V.A.\u00a0Sindagi and V.M.\u00a0Patel, MeGA-CDA: Memory guided attention for category-aware unsupervised domain adaptive object detection, in: Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2021, pp.\u00a04516\u20134526."},{"key":"10.3233\/AIC-220039_ref53","unstructured":"C.H.\u00a0Yu and J.\u00a0Dwang, Transfer learning with dynamic adversarial adaptation network, in: Proceedings of the 2019 IEEE Conference on Computer Vision and Pattern Recognition, IEEE, Piscataway, 2019, pp.\u00a0936\u2013944."},{"key":"10.3233\/AIC-220039_ref54","doi-asserted-by":"crossref","unstructured":"J.-Y.\u00a0Zhu, T.\u00a0Park, P.\u00a0Isola and A.\u00a0Efros, Unpaired image-to-image translation using cycle-consistent adversarial networks, in: Proceedings of the IEEE International Conference on Computer Vision, 2017, pp.\u00a02223\u20132232.","DOI":"10.1109\/ICCV.2017.244"},{"key":"10.3233\/AIC-220039_ref55","doi-asserted-by":"crossref","unstructured":"X.\u00a0Zhu, J.\u00a0Pang, C.\u00a0Yang, J.\u00a0Shi and D.\u00a0Lin, Adapting object detectors via selective cross-domain alignment, in: CVPR, 2019.","DOI":"10.1109\/CVPR.2019.00078"}],"container-title":["AI Communications"],"original-title":[],"link":[{"URL":"https:\/\/content.iospress.com\/download?id=10.3233\/AIC-220039","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,4,28]],"date-time":"2026-04-28T18:28:06Z","timestamp":1777400886000},"score":1,"resource":{"primary":{"URL":"https:\/\/journals.sagepub.com\/doi\/full\/10.3233\/AIC-220039"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,2,9]]},"references-count":46,"journal-issue":{"issue":"1"},"URL":"https:\/\/doi.org\/10.3233\/aic-220039","relation":{},"ISSN":["1875-8452","0921-7126"],"issn-type":[{"value":"1875-8452","type":"electronic"},{"value":"0921-7126","type":"print"}],"subject":[],"published":{"date-parts":[[2023,2,9]]}}}