{"type":"ScholarlyArticle","title":"Adversarial Constraint Learning for Structured Prediction","authors":[{"name":"Hongyu Ren","slug":"hongyu-ren","given":"Hongyu","family":"Ren","affiliation":"Department of Computer Science , Stanford University"},{"name":"Russell Stewart","slug":"russell-stewart","given":"Russell","family":"Stewart","affiliation":"Department of Computer Science , Stanford University"},{"name":"Jiaming Song","slug":"jiaming-song","given":"Jiaming","family":"Song","affiliation":"Department of Computer Science , Stanford University"},{"name":"Volodymyr Kuleshov","slug":"volodymyr-kuleshov","given":"Volodymyr","family":"Kuleshov","affiliation":"Department of Computer Science , Stanford University"},{"name":"Stefano Ermon","slug":"stefano-ermon","given":"Stefano","family":"Ermon","affiliation":"Department of Computer Science , Stanford University"}],"abstract":"Constraint-based learning reduces the burden of collecting labels by having users specify general properties of structured outputs, such as constraints imposed by physical laws. We propose a novel framework for simultaneously learning these constraints and using them for supervision, bypassing the difficulty of using domain expertise to manually specify constraints. Learning requires a black-box simulator of structured outputs, which generates valid labels, but need not model their corresponding inputs or the input-label relationship. At training time, we constrain the model to produce outputs that cannot be distinguished from simulated labels by adversarial training. Providing our framework with a small number of labeled inputs gives rise to a new semi-supervised structured prediction model; we evaluate this model on multiple tasks --- tracking, pose estimation and time series prediction --- and find that it achieves high accuracy with only a small number of labeled inputs. In some cases, no labels are required at all.","doi":"10.24963/ijcai.2018/366","url":"https://ijcai.org/proceedings/2018/366","pdf":"https://ijcai.org/proceedings/2018/0366.pdf","markdown":"https://ijcai.org/proceedings/2018/366.md","bibtex":"https://ijcai.org/proceedings/2018/bibtex/366","volume":{"slug":"2018","title":"Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2018","url":"https://ijcai.org/proceedings/2018/","year":2018,"location":"Stockholm","dates":"13-19 July 2018","editors":"Jérôme Lang","isbn":"978-0-9992411-2-7","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2018"},"section":"Machine Learning","track":"Main track","number":366,"pages":{"first":2637,"last":2643},"keywords":["Machine Learning: Semi-Supervised Learning","Machine Learning: Structured Prediction"],"videos":[],"references":[{"position":1,"raw":"[Arjovsky et al., 2017] Martin Arjovsky, Soumith Chintala, and Léon Bottou. Wasserstein gan. arXiv preprint arXiv:1701.07875, 2017.","title":"Wasserstein gan","authors":"Martin Arjovsky, Soumith Chintala, Léon Bottou","year":"2017","venue":null,"doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[Chang et al., 2007] Ming-Wei Chang, Lev Ratinov, and Dan Roth. Guiding semi-supervision with constraint- driven learning. In ACL, pages 280-287, 2007.","title":"Guiding semi-supervision with constraintdriven learning","authors":"Ming-Wei Chang, Lev Ratinov, Dan Roth","year":"2007","venue":"ACL","doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[Choi et al., 2015] Arthur Choi, Guy Van den Broeck, and Adnan Darwiche. Tractable learning for structured proba- bility spaces: A case study in learning preference distribu- tions. In Proceedings of 24th International Joint Confer- ence on Artificial Intelligence (IJCAI), 2015.","title":"Tractable learning for structured probability spaces: A case study in learning preference distributions","authors":"Arthur Choi, Guy Van Den, Adnan Broeck, Darwiche","year":"2015","venue":"Proceedings of 24th International Joint Conference on Artificial Intelligence (IJCAI)","doi":null,"url":null,"internal_url":"/proceedings/2015/405"},{"position":4,"raw":"[Ermon et al., 2012] Stefano Ermon, Ronan Le Bras, Carla P Gomes, Bart Selman, and R Bruce Van Dover. Smt-aided combinatorial materials discovery. In International Con- ference on Theory and Applications of Satisfiability Test- ing, pages 172-185. Springer, 2012.","title":"Smt-aided combinatorial materials discovery","authors":"Stefano Ermon, Le Ronan, Carla P Bras, Bart Gomes, R Selman, Van Bruce, Dover","year":"2012","venue":"International Conference on Theory and Applications of Satisfiability Testing","doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[Ermon et al., 2015] Stefano Ermon, Ronan Le Bras, San- tosh K Suram, John M Gregoire, Carla P Gomes, Bart Selman, and Robert Bruce van Dover. Pattern decomposi- tion with complex combinatorial constraints: Application to materials discovery. In AAAI, pages 636-643, 2015.","title":"Pattern decomposition with complex combinatorial constraints: Application to materials discovery","authors":"Stefano Ermon, Le Ronan, Bras, K Santosh, John M Suram, Carla P Gregoire, Bart Gomes, Robert Selman, Bruce Van Dover","year":"2015","venue":"AAAI","doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[Goodfellow et al., 2014] Ian Goodfellow, Jean Pouget- Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. Gen- erative adversarial nets. In Advances in neural information processing systems, pages 2672-2680, 2014.","title":"Generative adversarial nets","authors":"Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, Yoshua Bengio","year":"2014","venue":"Advances in neural information processing systems","doi":null,"url":null,"internal_url":null},{"position":7,"raw":"[Gulrajani et al., 2017] Ishaan Gulrajani, Faruk Ahmed, Martin Arjovsky, Vincent Dumoulin, and Aaron Courville. Improved training of wasserstein gans. arXiv preprint arXiv:1704.00028, 2017.","title":"Improved training of wasserstein gans","authors":"Ishaan Gulrajani, Faruk Ahmed, Martin Arjovsky, Aaron Vincent Dumoulin, Courville","year":"2017","venue":null,"doi":null,"url":null,"internal_url":null},{"position":8,"raw":"[Huang et al., 2014] Dong Huang, Shitong Yao, Yi Wang, and Fernando De La Torre. Sequential max-margin event detectors. In European conference on computer vision, pages 410-424. Springer, 2014.","title":"Sequential max-margin event detectors","authors":"Dong Huang, Shitong Yao, Yi Wang, Fernando De, La Torre","year":"2014","venue":"European conference on computer vision","doi":null,"url":null,"internal_url":null},{"position":9,"raw":"[Kingma et al., 2014] Diederik P Kingma, Shakir Mohamed, Danilo Jimenez Rezende, and Max Welling. Semi- supervised learning with deep generative models. In Ad- vances in Neural Information Processing Systems, pages 3581-3589, 2014.","title":"Semisupervised learning with deep generative models","authors":"Shakir Diederik P Kingma, Danilo Mohamed, Max Jimenez Rezende, Welling","year":"2014","venue":"Advances in Neural Information Processing Systems","doi":null,"url":null,"internal_url":null},{"position":10,"raw":"[Koller and Friedman, 2009] Daphne Koller and Nir Fried- man. Probabilistic graphical models: principles and tech- niques. MIT press, 2009.","title":"Probabilistic graphical models: principles and techniques","authors":"Friedman Koller","year":"2009","venue":null,"doi":null,"url":null,"internal_url":null},{"position":11,"raw":"[Li et al., 2016] Chongxuan Li, Jun Zhu, and Bo Zhang. Max-margin deep generative models for (semi-) super- vised learning. arXiv preprint arXiv:1611.07119, 2016.","title":"Max-margin deep generative models for (semi-) supervised learning","authors":"Li","year":"2016","venue":null,"doi":null,"url":null,"internal_url":null},{"position":12,"raw":"[Li et al., 2017] Yunzhu Li, Jiaming Song, and Stefano Er- mon. Infogail: Interpretable imitation learning from vi- sual demonstrations. In Advances in Neural Information Processing Systems, pages 3815-3825, 2017.","title":"Ermon. Infogail: Interpretable imitation learning from visual demonstrations","authors":"Li","year":"2017","venue":"Advances in Neural Information Processing Systems","doi":null,"url":null,"internal_url":null},{"position":13,"raw":"[Miyato et al., 2017] Takeru Miyato, Shin-ichi Maeda, Masanori Koyama, and Shin Ishii. Virtual adversarial training: a regularization method for supervised and semi- supervised learning. arXiv preprint arXiv:1704.03976, 2017.","title":"Virtual adversarial training: a regularization method for supervised and semisupervised learning","authors":"Miyato","year":"2017","venue":null,"doi":null,"url":null,"internal_url":null},{"position":14,"raw":"[Mohamed and Lakshminarayanan, 2016] Shakir Mohamed and Balaji Lakshminarayanan. Learning in implicit gen- erative models. arXiv preprint arXiv:1610.03483, 2016.","title":"Learning in implicit generative models","authors":"Shakir Mohamed, Balaji Lakshminarayanan","year":"2016","venue":null,"doi":null,"url":null,"internal_url":null},{"position":15,"raw":"[Papandreou et al., 2015] George Papandreou, Liang-Chieh Chen, Kevin Murphy, and Alan L Yuille. Weakly-and semi-supervised learning of a dcnn for semantic image segmentation. arXiv preprint arXiv:1502.02734, 2015.","title":"Weakly-and semi-supervised learning of a dcnn for semantic image segmentation","authors":"George Papandreou, Liang-Chieh Chen, Kevin Murphy, Alan L Yuille","year":"2015","venue":null,"doi":null,"url":null,"internal_url":null},{"position":16,"raw":"[Pathak et al., 2015] Deepak Pathak, Philipp Krahenbuhl, and Trevor Darrell. Constrained convolutional neural net- works for weakly supervised segmentation. In Proceed- ings of the IEEE international conference on computer vi- sion, pages 1796-1804, 2015.","title":"Constrained convolutional neural networks for weakly supervised segmentation","authors":"Deepak Pathak, Philipp Krahenbuhl, Trevor Darrell","year":"2015","venue":"Proceedings of the IEEE international conference on computer vision","doi":null,"url":null,"internal_url":null},{"position":17,"raw":"[Radford et al., 2015] Alec Radford, Luke Metz, and Soumith Chintala. Unsupervised representation learning with deep convolutional generative adversarial networks. arXiv preprint arXiv:1511.06434, 2015.","title":"Unsupervised representation learning with deep convolutional generative adversarial networks","authors":"Radford","year":"2015","venue":null,"doi":null,"url":null,"internal_url":null},{"position":18,"raw":"[Ratner et al., 2016] Alexander J Ratner, Christopher M De Sa, Sen Wu, Daniel Selsam, and Christopher Ré. Data programming: Creating large training sets, quickly. In Ad- vances in Neural Information Processing Systems, pages 3567-3575, 2016.","title":"Data programming: Creating large training sets, quickly","authors":"Christopher M De Alexander J Ratner, Sen Sa, Daniel Wu, Christopher Selsam, Ré","year":"2016","venue":"Advances in Neural Information Processing Systems","doi":null,"url":null,"internal_url":null},{"position":19,"raw":"[Richardson and Domingos, 2006] Matthew Richardson and Pedro Domingos. Markov logic networks. Machine learn- ing, 62(1):107-136, 2006.","title":"Matthew Richardson and Pedro Domingos. Markov logic networks","authors":"Domingos Richardson","year":"2006","venue":"Machine learning","doi":null,"url":null,"internal_url":null},{"position":20,"raw":"[Salimans et al., 2016] Tim Salimans, Ian Goodfellow, Woj- ciech Zaremba, Vicki Cheung, Alec Radford, and Xi Chen. Improved techniques for training gans. In Advances in Neural Information Processing Systems, pages 2234- 2242, 2016.","title":"Improved techniques for training gans","authors":"Tim Salimans, Ian Goodfellow, Wojciech Zaremba, Vicki Cheung, Alec Radford, Xi Chen","year":"2016","venue":"Advances in Neural Information Processing Systems","doi":null,"url":null,"internal_url":null},{"position":21,"raw":"[Shcherbatyi and Andres, 2016] Iaroslav Shcherbatyi and Bjoern Andres. Convexification of learning from con- straints. In German Conference on Pattern Recognition, pages 79-90. Springer, 2016.","title":"Iaroslav Shcherbatyi and Bjoern Andres. Convexification of learning from constraints","authors":"Andres Shcherbatyi","year":"2016","venue":"German Conference on Pattern Recognition","doi":null,"url":null,"internal_url":null},{"position":22,"raw":"[Stewart and Ermon, 2017] Russell Stewart and Stefano Er- mon. Label-free supervision of neural networks with physics and domain knowledge. In AAAI, pages 2576- 2582, 2017.","title":"Label-free supervision of neural networks with physics and domain knowledge","authors":"Russell Stewart, Stefano Ermon","year":"2017","venue":"AAAI","doi":null,"url":null,"internal_url":null},{"position":23,"raw":"[Xu et al., 2017] Jingyi Xu, Zilu Zhang, Tal Friedman, Yitao Liang, and Guy Van den Broeck. A semantic loss function for deep learning with symbolic knowledge. arXiv preprint arXiv:1711.11157, 2017.","title":"A semantic loss function for deep learning with symbolic knowledge","authors":"Jingyi Xu, Zilu Zhang, Tal Friedman, Yitao Liang, Guy Van Den Broeck","year":"2017","venue":null,"doi":null,"url":null,"internal_url":null},{"position":24,"raw":"[Yang and Ramanan, 2013] Yi Yang and Deva Ramanan. Ar- ticulated human detection with flexible mixtures of parts. IEEE Transactions on Pattern Analysis and Machine In- telligence, 35(12):2878-2890, 2013.","title":"Articulated human detection with flexible mixtures of parts","authors":"Yang, Ramanan, Yi Yang, Deva Ramanan","year":"2013","venue":"IEEE Transactions on Pattern Analysis and Machine Intelligence","doi":null,"url":null,"internal_url":null},{"position":25,"raw":"[Zamora-Martínez et al., 2014] F Zamora-Martínez, P Romeu, P Botella-Rocamora, and J Pardo. On- line learning of indoor temperature forecasting models towards energy efficiency. Energy and Buildings, 83:162-172, 2014.","title":"Online learning of indoor temperature forecasting models towards energy efficiency","authors":"Zamora-Martínez","year":"2014","venue":"Energy and Buildings","doi":null,"url":null,"internal_url":null}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}