{"type":"ScholarlyArticle","title":"Trajectory Similarity Learning with Auxiliary Supervision and Optimal Matching","authors":[{"name":"Hanyuan Zhang","slug":"hanyuan-zhang","given":"Hanyuan","family":"Zhang","affiliation":"School of Computer Science , Fudan University; Shanghai Key Laboratory of Data Science , Fudan University; Shanghai Institute of Intelligent Electronics & Systems"},{"name":"Xinyu Zhang","slug":"xinyu-zhang","given":"Xinyu","family":"Zhang","affiliation":"School of Computer Science , Fudan University; Shanghai Key Laboratory of Data Science , Fudan University; Shanghai Institute of Intelligent Electronics & Systems"},{"name":"Qize Jiang","slug":"qize-jiang","given":"Qize","family":"Jiang","affiliation":"School of Computer Science , Fudan University; Shanghai Key Laboratory of Data Science , Fudan University; Shanghai Institute of Intelligent Electronics & Systems"},{"name":"Baihua Zheng","slug":"baihua-zheng","given":"Baihua","family":"Zheng","affiliation":"Singapore Management University"},{"name":"Zhenbang Sun","slug":"zhenbang-sun","given":"Zhenbang","family":"Sun","affiliation":"ByteDance AI Lab , Beijing , China"},{"name":"Weiwei Sun","slug":"weiwei-sun","given":"Weiwei","family":"Sun","affiliation":"School of Computer Science , Fudan University; Shanghai Key Laboratory of Data Science , Fudan University; Shanghai Institute of Intelligent Electronics & Systems"},{"name":"Changhu Wang","slug":"changhu-wang","given":"Changhu","family":"Wang","affiliation":"ByteDance AI Lab , Beijing , China"}],"abstract":"Trajectory similarity computation is a core problem in the field of trajectory data queries. However, the high time complexity of calculating the trajectory similarity has always been a bottleneck in real-world applications. Learning-based methods can map trajectories into a uniform embedding space to calculate the similarity of two trajectories with embeddings in constant time. In this paper, we propose a novel trajectory representation learning framework Traj2SimVec that performs scalable and robust trajectory similarity computation. We use a simple and fast trajectory simplification and indexing approach to obtain triplet training samples efficiently. We make the framework more robust via taking full use of the sub-trajectory similarity information as auxiliary supervision. Furthermore, the framework supports the point matching query by modeling the optimal matching relationship of trajectory points under different distance metrics. The comprehensive experiments on real-world datasets demonstrate that our model substantially outperforms all existing approaches.","doi":"10.24963/ijcai.2020/444","url":"https://ijcai.org/proceedings/2020/444","pdf":"https://ijcai.org/proceedings/2020/0444.pdf","markdown":"https://ijcai.org/proceedings/2020/444.md","bibtex":"https://ijcai.org/proceedings/2020/bibtex/444","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":444,"pages":{"first":3209,"last":3215},"keywords":["Data Mining: Mining Spatial, Temporal Data","Machine Learning: Deep Learning","Multidisciplinary Topics and Applications: Transportation"],"videos":[{"label":"Video","url":"https://videos.ijcai20.org/24571/file_24571.m3u8","length_sec":300,"provider":"hls"}],"references":[{"position":1,"raw":"[Agarwal et al., 2015] Pankaj K Agarwal, Kyle Fox, Jiang- wei Pan, and Rex Ying. Approximating dynamic time warping and edit distance for a pair of point sequences. arXiv preprint arXiv:1512.01876, 2015.","title":"Approximating dynamic time warping and edit distance for a pair of point sequences","authors":"Kyle Pankaj K Agarwal, Jiangwei Fox, Rex Pan, Ying","year":"2015","venue":null,"doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[Bellman and others, 1954] Richard Bellman et al. The the- ory of dynamic programming. Bulletin of the AMS, 60(6):503-515, 1954.","title":"The theory of dynamic programming","authors":"Richard Bellman","year":"1954","venue":"Bulletin of the AMS","doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[Belogay et al., 1997] E Belogay, C Cabrelli, U Molter, and R Shonkwiler. Calculating the hausdorff distance between curves. Information Processing Letters, 64, 1997.","title":"Calculating the hausdorff distance between curves","authors":"Belogay, U Cabrelli, Molter, Shonkwiler","year":"1997","venue":"Information Processing Letters","doi":null,"url":null,"internal_url":null},{"position":4,"raw":"[Bentley, 1975] Jon Louis Bentley. Multidimensional bi- nary search trees used for associative searching. CACM, 18(9):509-517, 1975.","title":"Multidimensional binary search trees used for associative searching","authors":"Jon Louis Bentley, Bentley","year":"1975","venue":"CACM","doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[Chao et al., 2019] Pingfu Chao, Wen Hua, and Xiaofang Zhou. Trajectories know where map is wrong: an itera- tive framework for map-trajectory co-optimisation. World Wide Web, pages 1-27, 2019.","title":"Trajectories know where map is wrong: an iterative framework for map-trajectory co-optimisation","authors":"Chao","year":"2019","venue":"World Wide Web","doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[Chen et al., 2005] Lei Chen, M Tamer Özsu, and Vincent Oria. Robust and fast similarity search for moving object trajectories. In SIGMOD'05, pages 491-502, 2005.","title":"Robust and fast similarity search for moving object trajectories","authors":"Lei Chen, M Tamer Özsu, Vincent Oria","year":"2005","venue":"SIGMOD'05","doi":null,"url":null,"internal_url":null},{"position":7,"raw":"[Chen et al., 2009] Yukun Chen, Kai Jiang, Yu Zheng, Chunping Li, and Nenghai Yu. Trajectory simplification method for location-based social networking services. In LSBN'09, pages 33-40, 2009.","title":"Trajectory simplification method for location-based social networking services","authors":"Yukun Chen, Kai Jiang, Yu Zheng, Chunping Li, Nenghai Yu","year":"2009","venue":"LSBN'09","doi":null,"url":null,"internal_url":null},{"position":8,"raw":"[Chen et al., 2010] Zaiben Chen, Heng Tao Shen, Xiaofang Zhou, Yu Zheng, and Xing Xie. Searching trajectories by locations: an efficiency study. In SIGMOD'10, pages 255- 266, 2010.","title":"Searching trajectories by locations: an efficiency study","authors":"Zaiben Chen, Xiaofang Heng Tao Shen, Yu Zhou, Xing Zheng, Xie","year":"2010","venue":"SIGMOD'10","doi":null,"url":null,"internal_url":null},{"position":9,"raw":"[Chow et al., 2018] Ka-Ho Chow, Anish Hiranandani, Yifeng Zhang, and S-H Gary Chan. Representation learning of pedestrian trajectories using actor-critic sequence-to-sequence autoencoder. arXiv preprint arXiv:1811.08069, 2018.","title":"Representation learning of pedestrian trajectories using actor-critic sequence-to-sequence autoencoder","authors":"Ka-Ho Chow, Anish Hiranandani, Yifeng Zhang, S-H Gary, Chan","year":"2018","venue":null,"doi":null,"url":null,"internal_url":null},{"position":10,"raw":"[Chung et al., 2014] Junyoung Chung, Caglar Gulcehre, KyungHyun Cho, and Yoshua Bengio. Empirical evalua- tion of gated recurrent neural networks on sequence mod- eling. arXiv preprint arXiv:1412.3555, 2014.","title":"Empirical evaluation of gated recurrent neural networks on sequence modeling","authors":"Chung","year":"2014","venue":null,"doi":null,"url":null,"internal_url":null},{"position":11,"raw":"[Driemel and Silvestri, 2017] Anne Driemel and Francesco Silvestri. Locality-sensitive hashing of curves. arXiv preprint arXiv:1703.04040, 2017.","title":"Anne Driemel and Francesco Silvestri","authors":"Silvestri Driemel","year":"2017","venue":null,"doi":null,"url":null,"internal_url":null},{"position":12,"raw":"[Eiter and Mannila, 1994] Thomas Eiter and Heikki Man- nila. Computing discrete fréchet distance. Technical re- port, Citeseer, 1994.","title":"Computing discrete fréchet distance","authors":"Thomas Eiter, Heikki Mannila","year":"1994","venue":null,"doi":null,"url":null,"internal_url":null},{"position":13,"raw":"[Faghri et al., 2017] Fartash Faghri, David J Fleet, Jamie Ryan Kiros, and Sanja Fidler. Vse++: Im- proving visual-semantic embeddings with hard negatives. arXiv preprint arXiv:1707.05612, 2017.","title":"Vse++: Improving visual-semantic embeddings with hard negatives","authors":"Faghri","year":"2017","venue":null,"doi":null,"url":null,"internal_url":null},{"position":14,"raw":"[Gers et al., 1999] Felix A Gers, Jürgen Schmidhuber, and Fred Cummins. Learning to forget: Continual prediction with lstm. 1999.","title":"Learning to forget: Continual prediction with lstm","authors":"Jürgen Felix A Gers, Fred Schmidhuber, Cummins","year":"1999","venue":null,"doi":null,"url":null,"internal_url":null},{"position":15,"raw":"[Gowanlock and Casanova, 2015] Michael Gowanlock and Henri Casanova. Distance threshold similarity searches: Efficient trajectory indexing on the gpu. TPDS, 27(9):2533-2545, 2015.","title":"Distance threshold similarity searches: Efficient trajectory indexing on the gpu","authors":"Michael Gowanlock, Henri Casanova","year":"2015","venue":"TPDS","doi":null,"url":null,"internal_url":null},{"position":16,"raw":"[Kingma and Ba, 2014] Diederik P Kingma and Jimmy Ba. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980, 2014.","title":"Adam: A method for stochastic optimization","authors":"P Diederik, Jimmy Kingma, Ba","year":"2014","venue":null,"doi":null,"url":null,"internal_url":null},{"position":17,"raw":"[Lee et al., 2007] Jae-Gil Lee, Jiawei Han, and Kyu-Young Whang. Trajectory clustering: a partition-and-group framework. In SIGMOD'07, pages 593-604, 2007.","title":"Trajectory clustering: a partition-and-group framework","authors":"Jae-Gil Lee, Jiawei Han, Kyu-Young Whang","year":"2007","venue":"SIGMOD'07","doi":null,"url":null,"internal_url":null},{"position":18,"raw":"[Li et al., 2018] Xiucheng Li, Kaiqi Zhao, Gao Cong, Chris- tian S Jensen, and Wei Wei. Deep representation learning for trajectory similarity computation. In ICDE'18, pages 617-628, 2018.","title":"Deep representation learning for trajectory similarity computation","authors":"Li","year":"2018","venue":"ICDE'18","doi":null,"url":null,"internal_url":null},{"position":19,"raw":"[Meng et al., 2019] Fanrong Meng, Guan Yuan, Shaoqian Lv, Zhixiao Wang, and Shixiong Xia. An overview on trajectory outlier detection. Artificial Intelligence Review, 52(4):2437-2456, 2019.","title":"An overview on trajectory outlier detection","authors":"Fanrong Meng, Guan Yuan, Shaoqian Lv, Zhixiao Wang, Shixiong Xia","year":"2019","venue":"Artificial Intelligence Review","doi":null,"url":null,"internal_url":null},{"position":20,"raw":"[Müller, 2007] Meinard Müller. Dynamic time warping. In- formation retrieval for music and motion, pages 69-84, 2007.","title":"Dynamic time warping. Information retrieval for music and motion","authors":"Meinard Müller, Müller","year":"2007","venue":null,"doi":null,"url":null,"internal_url":null},{"position":21,"raw":"[Schroff et al., 2015] Florian Schroff, Dmitry Kalenichenko, and James Philbin. Facenet: A unified embedding for face recognition and clustering. In CVPR'15, pages 815-823, 2015.","title":"Facenet: A unified embedding for face recognition and clustering","authors":"Schroff","year":"2015","venue":"CVPR'15","doi":null,"url":null,"internal_url":null},{"position":22,"raw":"[Shan et al., 2015] Zhangqing Shan, Hao Wu, Weiwei Sun, and Baihua Zheng. Cobweb: a robust map update system using gps trajectories. In UBICOMP'15, pages 927-937, 2015.","title":"Cobweb: a robust map update system using gps trajectories","authors":"Shan","year":"2015","venue":"UBICOMP'","doi":null,"url":null,"internal_url":null},{"position":23,"raw":"[Shang et al., 2017] Shuo Shang, Lisi Chen, Zhewei Wei, Christian S Jensen, Kai Zheng, and Panos Kalnis. Tra- jectory similarity join in spatial networks. Proceedings of the VLDB Endowment, 10(11):1178-1189, 2017.","title":"Trajectory similarity join in spatial networks","authors":"Shuo Shang, Lisi Chen, Zhewei Wei, Christian S Jensen, Kai Zheng, Panos Kalnis","year":"2017","venue":"Proceedings of the VLDB Endowment","doi":null,"url":null,"internal_url":null},{"position":24,"raw":"[Wang et al., 2019] Zheng Wang, Cheng Long, Gao Cong, and Ce Ju. Effective and efficient sports play retrieval with deep representation learning. In SIGKDD'19, pages 499- 509, 2019.","title":"Effective and efficient sports play retrieval with deep representation learning","authors":"Zheng Wang, Cheng Long, Gao Cong, Ce Ju","year":"2019","venue":"SIGKDD'19","doi":null,"url":null,"internal_url":null},{"position":25,"raw":"[Wu et al., 2017] Hao Wu, Weiwei Sun, and Baihua Zheng. A fast trajectory outlier detection approach via driving be- havior modeling. In CIKM'17, pages 837-846, 2017.","title":"A fast trajectory outlier detection approach via driving behavior modeling","authors":"Hao Wu, Weiwei Sun, Baihua Zheng","year":"2017","venue":"CIKM'17","doi":null,"url":null,"internal_url":null},{"position":26,"raw":"[Yao et al., 2018] Di Yao, Chao Zhang, Zhihua Zhu, Qin Hu, Zheng Wang, Jianhui Huang, and Jingping Bi. Learning deep representation for trajectory clustering. Expert Sys- tems, 35(2):e12252, 2018.","title":"Learning deep representation for trajectory clustering","authors":"Di Yao, Chao Zhang, Zhihua Zhu, Qin Hu, Zheng Wang, Jianhui Huang, Jingping Bi","year":"2018","venue":"Expert Systems","doi":null,"url":null,"internal_url":null},{"position":27,"raw":"[Yao et al., 2019] Di Yao, Gao Cong, Chao Zhang, and Jing- ping Bi. Computing trajectory similarity in linear time: A generic seed-guided neural metric learning approach. In ICDE'19, pages 1358-1369, 2019.","title":"Computing trajectory similarity in linear time: A generic seed-guided neural metric learning approach","authors":"Di Yao, Gao Cong, Chao Zhang, Jingping Bi","year":"2019","venue":"ICDE'19","doi":null,"url":null,"internal_url":null},{"position":28,"raw":"[Zhang et al., 2019] Yifan Zhang, An Liu, Guanfeng Liu, Zhixu Li, and Qing Li. Deep representation learning of ac- tivity trajectory similarity computation. In ICWS'19, pages 312-319, 2019.","title":"Deep representation learning of activity trajectory similarity computation","authors":"Yifan Zhang, An Liu, Guanfeng Liu, Zhixu Li, Qing Li","year":"2019","venue":"ICWS'19","doi":null,"url":null,"internal_url":null}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}