{"type":"ScholarlyArticle","title":"Clarinet: A One-step Approach Towards Budget-friendly Unsupervised Domain Adaptation","authors":[{"name":"Yiyang Zhang","slug":"yiyang-zhang","given":"Yiyang","family":"Zhang","affiliation":"Shenzhen International Graduate School , Tsinghua University; Centre for Artificial Intelligence , University of Technology Sydney"},{"name":"Feng Liu","slug":"feng-liu","given":"Feng","family":"Liu","affiliation":"Centre for Artificial Intelligence , University of Technology Sydney"},{"name":"Zhen Fang","slug":"zhen-fang","given":"Zhen","family":"Fang","affiliation":"Centre for Artificial Intelligence , University of Technology Sydney"},{"name":"Bo Yuan","slug":"bo-yuan","given":"Bo","family":"Yuan","affiliation":"Shenzhen International Graduate School , Tsinghua University"},{"name":"Guangquan Zhang","slug":"guangquan-zhang","given":"Guangquan","family":"Zhang","affiliation":"Centre for Artificial Intelligence , University of Technology Sydney"},{"name":"Jie Lu","slug":"jie-lu","given":"Jie","family":"Lu","affiliation":"Centre for Artificial Intelligence , University of Technology Sydney"}],"abstract":"In unsupervised domain adaptation (UDA), classifiers for the target domain are trained with massive true-label data from the source domain and unlabeled data from the target domain. However, it may be difficult to collect fully-true-label data in a source domain given limited budget. To mitigate this problem, we consider a novel problem setting where the classifier for the target domain has to be trained with complementary-label data from the source domain and unlabeled data from the target domain named budget-friendly UDA  (BFUDA). The key benefit is that it is much less costly to collect complementary-label source data (required by BFUDA) than collecting the true-label source data (required by ordinary UDA). To this end, complementary label adversarial network (CLARINET) is proposed to solve the BFUDA problem. CLARINET maintains two deep networks simultaneously, where one focuses on classifying complementary-label source data and the other takes care of the source-to-target distributional adaptation. Experiments show that CLARINET significantly outperforms a series of competent baselines.","doi":"10.24963/ijcai.2020/350","url":"https://ijcai.org/proceedings/2020/350","pdf":"https://ijcai.org/proceedings/2020/0350.pdf","markdown":"https://ijcai.org/proceedings/2020/350.md","bibtex":"https://ijcai.org/proceedings/2020/bibtex/350","volume":{"slug":"2020","title":"Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2020","url":"https://ijcai.org/proceedings/2020/","year":2020,"location":"Yokohama","dates":"January 2021 (postponed from 11-17 July 2020)","editors":"Christian Bessiere","isbn":"978-0-9992411-6-5","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2020"},"section":"Machine Learning","track":"Main track","number":350,"pages":{"first":2526,"last":2532},"keywords":["Machine Learning: Semi-Supervised Learning","Machine Learning: Transfer, Adaptation, Multi-task Learning","Machine Learning: Deep Learning"],"videos":[{"label":"Video","url":"https://videos.ijcai20.org/24495/file_24495.m3u8","length_sec":294,"provider":"hls"}],"references":[{"position":1,"raw":"[Agresti et al., 2019] G. Agresti, H. Schäfer, P. Sartor, and P. Zanuttigh. Unsupervised domain adaptation for tof data denoising with adversarial learning. In CVPR, pages 5584-5593, 2019.","title":"Unsupervised domain adaptation for tof data denoising with adversarial learning","authors":"G Agresti, H Schäfer, P Sartor, P Zanuttigh","year":"2019","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[Ao et al., 2017] S. Ao, X. Li, and C.X. Ling. Fast general- ized distillation for semi-supervised domain adaptation. In AAAI, pages 1719-1725, 2017.","title":"Fast generalized distillation for semi-supervised domain adaptation","authors":"S Ao, X Li, C X Ling","year":"2017","venue":"AAAI","doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[Ben-David et al., 2010] S. Ben-David, J. Blitzer, K. Cram- mer, A. Kulesza, F. Pereira, and J.W. Vaughan. A theory of learning from different domains. MLJ, 79(1-2):151-175, 2010.","title":"A theory of learning from different domains","authors":"S Ben-David, J Blitzer, K Crammer, A Kulesza, F Pereira, J W Vaughan","year":"2010","venue":"MLJ","doi":null,"url":null,"internal_url":null},{"position":4,"raw":"[Deng et al., 2019] Z. Deng, Y. Luo, and J. Zhu. Cluster alignment with a teacher for unsupervised domain adap- tation. In ICCV, pages 9944-9953, 2019.","title":"Cluster alignment with a teacher for unsupervised domain adaptation","authors":"Z Deng, Y Luo, J Zhu","year":"2019","venue":"ICCV","doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[Fang et al., 2019] Zhen Fang, Jie Lu, Feng Liu, and Guangquan Zhang. Unsupervised domain adaptation with sphere retracting transformation. In IJCNN, pages 1-8. IEEE, 2019.","title":"Unsupervised domain adaptation with sphere retracting transformation","authors":"Zhen Fang, Jie Lu, Feng Liu, Guangquan Zhang","year":"2019","venue":"IJCNN","doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[Ganin et al., 2016] Y. Ganin, E. Ustinova, H. Ajakan, P. Germain, H. Larochelle, F. Laviolette, M. Marchand, and V. Lempitsky. Domain-adversarial training of neural networks. JMLR, 17(59):1-35, 2016.","title":"Domain-adversarial training of neural networks","authors":"Y Ganin, E Ustinova, H Ajakan, P Germain, H Larochelle, F Laviolette, M Marchand, V Lempitsky","year":"2016","venue":"JMLR","doi":null,"url":null,"internal_url":null},{"position":7,"raw":"[Gong et al., 2016] M. Gong, K. Zhang, T. Liu, D. Tao, C. Glymour, and B. Schölkopf. Domain adaptation with conditional transferable components. In ICML, pages 2839-2848, 2016.","title":"Domain adaptation with conditional transferable components","authors":"M Gong, K Zhang, T Liu, D Tao, C Glymour, B Schölkopf","year":"2016","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":8,"raw":"[Gong et al., 2018] M. Gong, K. Zhang, B. Huang, C. Gly- mour, D. Tao, and K. Batmanghelich. Causal genera- tive domain adaptation networks. CoRR, abs/1804.04333, 2018.","title":"Causal generative domain adaptation networks","authors":"M Gong, K Zhang, B Huang, C Glymour, D Tao, K Batmanghelich","year":"2018","venue":null,"doi":null,"url":null,"internal_url":null},{"position":9,"raw":"[Han et al., 2018a] B. Han, J. Yao, G. Niu, M. Zhou, I. Tsang, Y. Zhang, and M. Sugiyama. Masking: A new perspective of noisy supervision. In NeurIPS, pages 5836- 5846, 2018.","title":"Masking: A new perspective of noisy supervision","authors":"Han","year":"2018","venue":"NeurIPS","doi":null,"url":null,"internal_url":null},{"position":10,"raw":"[Han et al., 2018b] B. Han, Q. Yao, X. Yu, G. Niu, M. Xu, W. Hu, I. Tsang, and M. Sugiyama. Co-teaching: Robust training of deep neural networks with extremely noisy la- bels. In NeurIPS, pages 8527-8537, 2018.","title":"Co-teaching: Robust training of deep neural networks with extremely noisy labels","authors":"Han","year":"2018","venue":"NeurIPS","doi":null,"url":null,"internal_url":null},{"position":11,"raw":"[Ishida et al., 2017] T. Ishida, G. Niu, W. Hu, and M. Sugiyama. Learning from complementary labels. In NeurIPS, pages 5639-5649, 2017.","title":"Learning from complementary labels","authors":"T Ishida, G Niu, W Hu, M Sugiyama","year":"2017","venue":"NeurIPS","doi":null,"url":null,"internal_url":null},{"position":12,"raw":"[Ishida et al., 2019] T. Ishida, G. Niu, A. Menon, and M. Sugiyama. Complementary-label learning for arbitrary losses and models. In ICML, pages 2971-2980, 2019.","title":"Complementary-label learning for arbitrary losses and models","authors":"T Ishida, G Niu, A Menon, M Sugiyama","year":"2019","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":13,"raw":"[Kiryo et al., 2017] R. Kiryo, G. Niu, M.C. du Plessis, and M. Sugiyama. Positive-unlabeled learning with non- negative risk estimator. In NeurIPS, pages 1675-1685, 2017.","title":"Positive-unlabeled learning with nonnegative risk estimator","authors":"R Kiryo, G Niu, M C Du Plessis, M Sugiyama","year":"2017","venue":"NeurIPS","doi":null,"url":null,"internal_url":null},{"position":14,"raw":"[Kumar et al., 2017] A. Kumar, P. Sattigeri, and T. Fletcher. Semi-supervised learning with gans: Manifold invariance with improved inference. In NeurIPS, pages 5534-5544, 2017.","title":"Semi-supervised learning with gans: Manifold invariance with improved inference","authors":"A Kumar, P Sattigeri, T Fletcher","year":"2017","venue":"NeurIPS","doi":null,"url":null,"internal_url":null},{"position":15,"raw":"[Liu et al., 2019] F. Liu, J. Lu, B. Han, G. Niu, G. Zhang, and M. Sugiyama. Butterfly: A panacea for all difficul- ties in wildly unsupervised domain adaptation. In NeurIPS LTS Workshop, 2019.","title":"Butterfly: A panacea for all difficulties in wildly unsupervised domain adaptation","authors":"F Liu, J Lu, B Han, G Niu, G Zhang, M Sugiyama","year":"2019","venue":"NeurIPS LTS Workshop","doi":null,"url":null,"internal_url":null},{"position":16,"raw":"[Liu et al., 2020] F. Liu, G. Zhang, and J. Lu. Heterogeneous domain adaptation: An unsupervised approach. TNNLS, Early Access, 2020.","title":"Heterogeneous domain adaptation: An unsupervised approach","authors":"F Liu, G Zhang, J Lu","year":"2020","venue":"TNNLS, Early Access","doi":null,"url":null,"internal_url":null},{"position":17,"raw":"[Long et al., 2015] M. Long, Y. Cao, J. Wang, and M.I. Jor- dan. Learning transferable features with deep adaptation networks. In ICML, page 97-105, 2015.","title":"Learning transferable features with deep adaptation networks","authors":"M Long, Y Cao, J Wang, M I Jordan","year":"2015","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":18,"raw":"[Long et al., 2018] M. Long, Z. Cao, J. Wang, and M.I Jor- dan. Conditional adversarial domain adaptation. In NeurIPS, pages 1640-1650, 2018.","title":"Conditional adversarial domain adaptation","authors":"M Long, Z Cao, J Wang, M I Jordan","year":"2018","venue":"NeurIPS","doi":null,"url":null,"internal_url":null},{"position":19,"raw":"[Saito et al., 2017] K. Saito, Y. Ushiku, and T. Harada. Asymmetric tri-training for unsupervised domain adapta- tion. In ICML, pages 2988-2997, 2017.","title":"Asymmetric tri-training for unsupervised domain adaptation","authors":"K Saito, Y Ushiku, T Harada","year":"2017","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":20,"raw":"[Sakai et al., 2017] T. Sakai, M.C. du Plessis, G. Niu, and M. Sugiyama. Semi-supervised classification based on classification from positive and unlabeled data. In ICML, pages 2998-3006, 2017.","title":"Semi-supervised classification based on classification from positive and unlabeled data","authors":"T Sakai, M C Du Plessis, G Niu, M Sugiyama","year":"2017","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":21,"raw":"[Sankaranarayanan et al., 2018] S. Sankaranarayanan, Y. Balaji, C.D. Castillo, and R. Chellappa. Generate to adapt: Aligning domains using generative adversarial networks. In CVPR, pages 8503-8512, 2018.","title":"Generate to adapt: Aligning domains using generative adversarial networks","authors":"S Sankaranarayanan, Y Balaji, C D Castillo, R Chellappa","year":"2018","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":22,"raw":"[Shu et al., 2019] Y. Shu, Z. Cao, M. Long, and J. Wang. Transferable curriculum for weakly-supervised domain adaptation. In AAAI, pages 4951-4958, 2019.","title":"Transferable curriculum for weakly-supervised domain adaptation","authors":"Y Shu, Z Cao, M Long, J Wang","year":"2019","venue":"AAAI","doi":null,"url":null,"internal_url":null},{"position":23,"raw":"[Song et al., 2009] L. Song, J. Huang, A. Smola, and K. Fukumizu. Hilbert space embeddings of conditional distributions with applications to dynamical systems. In ICML, pages 961-968, 2009.","title":"Hilbert space embeddings of conditional distributions with applications to dynamical systems","authors":"L Song, J Huang, A Smola, K Fukumizu","year":"2009","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":24,"raw":"[Sukhija et al., 2016] S. Sukhija, N.C. Krishnan, and G. Singh. Supervised heterogeneous domain adaptation via random forests. In IJCAI, pages 2039-2045, 2016.","title":"Supervised heterogeneous domain adaptation via random forests","authors":"S Sukhija, N C Krishnan, G Singh","year":"2016","venue":"IJCAI","doi":null,"url":null,"internal_url":"/proceedings/2016/291"},{"position":25,"raw":"[Yan et al., 2017] Y. Yan, W. Li, M. K. P. Ng, M. Tan, H. Wu, H. Min, and Q. Wu. Learning discriminative correlation subspace for heterogeneous domain adaptation. In IJCAI, pages 3252-3258, 2017.","title":"Learning discriminative correlation subspace for heterogeneous domain adaptation","authors":"Y Yan, W Li, M K P Ng, M Tan, H Wu, H Min, Q Wu","year":"2017","venue":"IJCAI","doi":null,"url":null,"internal_url":"/proceedings/2017/454"},{"position":26,"raw":"[Yu et al., 2018] X. Yu, T. Liu, M. Gong, and D. Tao. Learn- ing with biased complementary labels. In ECCV, pages 69-85, 2018.","title":"Learning with biased complementary labels","authors":"Yu","year":"2018","venue":"ECCV","doi":null,"url":null,"internal_url":null},{"position":27,"raw":"[Zhang et al., 2015] K. Zhang, M. Gong, and B. Schölkopf. Multi-source domain adaptation: A causal view. In AAAI, pages 3150-3157, 2015.","title":"Multi-source domain adaptation: A causal view","authors":"K Zhang, M Gong, B Schölkopf","year":"2015","venue":"AAAI","doi":null,"url":null,"internal_url":null},{"position":28,"raw":"[Zhao et al., 2019] S. Zhao, H. Fu, M. Gong, and D. Tao. Geometry-aware symmetric domain adaptation for monocular depth estimation. In CVPR, pages 9788-9798, 2019.","title":"Geometry-aware symmetric domain adaptation for monocular depth estimation","authors":"S Zhao, H Fu, M Gong, D Tao","year":"2019","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":29,"raw":"[Zhou et al., 2019a] J. T. Zhou, S. J. Pan, and I. W. Tsang. A deep learning framework for hybrid heterogeneous trans- fer learning. AIJ, 275:310 -328, 2019.","title":"A deep learning framework for hybrid heterogeneous transfer learning","authors":"Zhou","year":"2019","venue":"AIJ","doi":null,"url":null,"internal_url":null},{"position":30,"raw":"[Zhou et al., 2019b] J. T. Zhou, H. Zhang, D. Jin, and X. Peng. Dual adversarial transfer for sequence labeling. TPAMI, Early Access, 2019.","title":"Dual adversarial transfer for sequence labeling","authors":"Zhou","year":"2019","venue":"TPAMI, Early Access","doi":null,"url":null,"internal_url":null}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}