{"type":"ScholarlyArticle","title":"Hi-Fi: Hierarchical Feature Integration for Skeleton Detection","authors":[{"name":"Kai Zhao","slug":"kai-zhao","given":"Kai","family":"Zhao","affiliation":"College of Computer and Control Engineering , Nankai University"},{"name":"Wei Shen","slug":"wei-shen","given":"Wei","family":"Shen","affiliation":"Key Laboratory of Specialty Fiber Optics and Optical Access Networks , Shanghai University"},{"name":"Shanghua Gao","slug":"shanghua-gao","given":"Shanghua","family":"Gao","affiliation":"College of Computer and Control Engineering , Nankai University"},{"name":"Dandan Li","slug":"dandan-li","given":"Dandan","family":"Li","affiliation":"Key Laboratory of Specialty Fiber Optics and Optical Access Networks , Shanghai University"},{"name":"Ming-Ming Cheng","slug":"ming-ming-cheng","given":"Ming-Ming","family":"Cheng","affiliation":"College of Computer and Control Engineering , Nankai University"}],"abstract":"In natural images, the scales (thickness) of object skeletons may\r\ndramatically vary among objects and object parts.\r\nThus, robust skeleton detection requires powerful multi-scale feature integration ability.\r\nTo address this issue, we present a new convolutional neural network (CNN) architecture\r\nby introducing a novel hierarchical feature integration mechanism,\r\nnamed Hi-Fi, to address the object skeleton detection problem.\r\nThe proposed CNN-based approach intrinsically captures high-level semantics from deeper layers,\r\nas well as low-level details from shallower layers.\r\nBy hierarchically integrating different CNN feature levels with bidirectional guidance,\r\nour approach (1) enables mutual refinement across features of different levels,\r\nand (2) possesses the strong ability to capture both rich object context\r\nand high-resolution details.\r\nExperimental results show that our method\r\nsignificantly outperforms the state-of-the-art methods\r\nin terms of effectively fusing features from very different scales,\r\nas evidenced by a considerable performance improvement on several benchmarks.","doi":"10.24963/ijcai.2018/166","url":"https://ijcai.org/proceedings/2018/166","pdf":"https://ijcai.org/proceedings/2018/0166.pdf","markdown":"https://ijcai.org/proceedings/2018/166.md","bibtex":"https://ijcai.org/proceedings/2018/bibtex/166","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":"Computer Vision","track":"Main track","number":166,"pages":{"first":1191,"last":1197},"keywords":["Computer Vision: Recognition: Detection, Categorization, Indexing, Matching, Retrieval, Semantic Interpretation","Computer Vision: Computer Vision"],"videos":[],"references":[{"position":1,"raw":"[Blum, 1967] H. Blum. Models for the perception of speech and visual forms. In A Transformation for extracting new descriptors of shape, chapter 10, pages 363-380. MIT Press, Boston, MA, USA, 1967.","title":"Models for the perception of speech and visual forms","authors":"H Blum","year":"1967","venue":"A Transformation for extracting new descriptors of shape","doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[Demirci et al., 2006] M Fatih Demirci, Ali Shokoufandeh, Yakov Keselman, Lars Bretzner, and Sven Dickinson. Ob- ject recognition as many-to-many feature matching. IJCV, 69(2):203-222, 2006.","title":"Object recognition as many-to-many feature matching","authors":"Ali Fatih Demirci, Yakov Shokoufandeh, Lars Keselman, Sven Bretzner, Dickinson","year":"2006","venue":"IJCV","doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[Girshick et al., 2011] Ross Girshick, Jamie Shotton, Push- meet Kohli, Antonio Criminisi, and Andrew Fitzgibbon. Efficient regression of general-activity human poses from depth images. In ICCV, pages 415-422. IEEE, 2011.","title":"Efficient regression of general-activity human poses from depth images","authors":"Ross Girshick, Jamie Shotton, Pushmeet Kohli, Antonio Criminisi, Andrew Fitzgibbon","year":"2011","venue":"ICCV","doi":null,"url":null,"internal_url":null},{"position":4,"raw":"[He et al., 2016] Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recog- nition. In CVPR, pages 770-778, 2016.","title":"Deep residual learning for image recognition","authors":"Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun","year":"2016","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[Hou et al., 2018] Qibin Hou, Ming-Ming Cheng, Xiaowei Hu, Ali Borji, Zhuowen Tu, and Philip Torr. Deeply su- pervised salient object detection with short connections. PAMI, 2018.","title":"Deeply supervised salient object detection with short connections","authors":"Hou","year":"2018","venue":"PAMI","doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[Jang and Hong, 2001] Jeong-Hun Jang and Ki-Sang Hong. A pseudo-distance map for the segmentation-free skele- tonization of gray-scale images. In ICCV, volume 2, pages 18-23. IEEE, 2001.","title":"A pseudo-distance map for the segmentation-free skeletonization of gray-scale images","authors":"Jeong-Hun Jang, Ki-Sang Hong","year":"2001","venue":"ICCV","doi":null,"url":null,"internal_url":null},{"position":7,"raw":"[Jia et al., 2014] Yangqing Jia, Evan Shelhamer, Jeff Don- ahue, Sergey Karayev, Jonathan Long, Ross Girshick, Ser- gio Guadarrama, and Trevor Darrell. Caffe: Convolutional architecture for fast feature embedding. In ACMMM, pages 675-678. ACM, 2014.","title":"Caffe: Convolutional architecture for fast feature embedding","authors":"Jia","year":"2014","venue":"ACMMM","doi":null,"url":null,"internal_url":null},{"position":8,"raw":"[Ke et al., 2017] Wei Ke, Jie Chen, Jianbin Jiao, Guoying Zhao, and Qixiang Ye. Srn: Side-output residual network for object symmetry detection in the wild. In CVPR, July 2017.","title":"Srn: Side-output residual network for object symmetry detection in the wild","authors":"Wei Ke, Jie Chen, Jianbin Jiao, Guoying Zhao, Qixiang Ye","year":"2017","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":9,"raw":"[Liu et al., 2017] Yun Liu, Ming-Ming Cheng, Xiaowei Hu, Kai Wang, and Xiang Bai. Richer convolutional features for edge detection. In CVPR, 2017.","title":"Richer convolutional features for edge detection","authors":"Yun Liu, Ming-Ming Cheng, Xiaowei Hu, Kai Wang, Xiang Bai","year":"2017","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":10,"raw":"[Morse et al., 1993] Bryan S Morse, Stephen M Pizer, and Alan Liu. Multiscale medial analysis of medical images. In IPMI, pages 112-131. Springer, 1993.","title":"Multiscale medial analysis of medical images","authors":"Stephen M Bryan S Morse, Alan Pizer, Liu","year":"1993","venue":"IPMI","doi":null,"url":null,"internal_url":null},{"position":11,"raw":"[Sebastian et al., 2004] Thomas B Sebastian, Philip N Klein, and Benjamin B Kimia. Recognition of shapes by editing their shock graphs. PAMI, 26(5):550-571, 2004.","title":"Recognition of shapes by editing their shock graphs","authors":"Philip N Thomas B Sebastian, Benjamin B Klein, Kimia","year":"2004","venue":"PAMI","doi":null,"url":null,"internal_url":null},{"position":12,"raw":"[Shen et al., 2016a] Wei Shen, Xiang Bai, Zihao Hu, and Zhijiang Zhang. Multiple instance subspace learning via partial random projection tree for local reflection symme- try in natural images. PR, 52:306-316, 2016.","title":"Multiple instance subspace learning via partial random projection tree for local reflection symmetry in natural images","authors":"Shen","year":"2016","venue":"PR","doi":null,"url":null,"internal_url":null},{"position":13,"raw":"[Shen et al., 2016b] Wei Shen, Kai Zhao, Yuan Jiang, Yan Wang, Zhijiang Zhang, and Xiang Bai. Object skeleton ex- traction in natural images by fusing scale-associated deep side outputs. CVPR, pages 222-230, 2016.","title":"Object skeleton extraction in natural images by fusing scale-associated deep side outputs","authors":"Shen","year":"2016","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":14,"raw":"[Shen et al., 2017] Wei Shen, Kai Zhao, Yuan Jiang, Yan Wang, Xiang Bai, and Alan Yuille. Deepskeleton: Learn- ing multi-task scale-associated deep side outputs for object skeleton extraction in natural images. TIP, 26(11):5298- 5311, 2017.","title":"Deepskeleton: Learning multi-task scale-associated deep side outputs for object skeleton extraction in natural images","authors":"Wei Shen, Kai Zhao, Yuan Jiang, Yan Wang, Xiang Bai, Alan Yuille","year":"2017","venue":"TIP","doi":null,"url":null,"internal_url":null},{"position":15,"raw":"[Shotton et al., 2013] Jamie Shotton, Toby Sharp, Alex Kip- man, Andrew Fitzgibbon, Mark Finocchio, Andrew Blake, Mat Cook, and Richard Moore. Real-time human pose recognition in parts from single depth images. CACM, 56(1):116-124, 2013.","title":"Real-time human pose recognition in parts from single depth images","authors":"Jamie Shotton, Toby Sharp, Alex Kipman, Andrew Fitzgibbon, Mark Finocchio, Andrew Blake, Mat Cook, Richard Moore","year":"2013","venue":"CACM","doi":null,"url":null,"internal_url":null},{"position":16,"raw":"[Simonyan and Zisserman, 2015] Karen Simonyan and An- drew Zisserman. Very deep convolutional networks for large-scale image recognition. In ICLR, 2015.","title":"Very deep convolutional networks for large-scale image recognition","authors":"Karen Simonyan, Andrew Zisserman","year":"2015","venue":"ICLR","doi":null,"url":null,"internal_url":null},{"position":17,"raw":"[Sironi et al., 2014] Amos Sironi, Vincent Lepetit, and Pas- cal Fua. Multiscale centerline detection by learning a scale-space distance transform. In CVPR, pages 2697- 2704, 2014.","title":"Multiscale centerline detection by learning a scale-space distance transform","authors":"Amos Sironi, Vincent Lepetit, Pascal Fua","year":"2014","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":18,"raw":"[Sun et al., 2012] Min Sun, Pushmeet Kohli, and Jamie Shotton. Conditional regression forests for human pose estimation. In CVPR, pages 3394-3401. IEEE, 2012.","title":"Conditional regression forests for human pose estimation","authors":"Sun","year":"2012","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":19,"raw":"[Tsogkas and Dickinson, 2017] Stavros Tsogkas and Sven Dickinson. Amat: Medial axis transform for natural im- ages. ICCV, 2017.","title":"Stavros Tsogkas and Sven Dickinson. Amat: Medial axis transform for natural images","authors":"Dickinson Tsogkas","year":"2017","venue":null,"doi":null,"url":null,"internal_url":null},{"position":20,"raw":"[Tsogkas and Kokkinos, 2012] Stavros Tsogkas and Iasonas Kokkinos. Learning-based symmetry detection in natural images. In ECCV, pages 41-54. Springer, 2012.","title":"Stavros Tsogkas and Iasonas Kokkinos. Learning-based symmetry detection in natural images","authors":"Kokkinos Tsogkas","year":"2012","venue":"ECCV","doi":null,"url":null,"internal_url":null},{"position":21,"raw":"[Xie and Tu, 2015] Saining Xie and Zhuowen Tu. Holistically-nested edge detection. In ICCV, pages 1395-1403, 2015.","title":"Saining Xie and Zhuowen Tu. Holistically-nested edge detection","authors":"Tu Xie","year":"2015","venue":"ICCV","doi":null,"url":null,"internal_url":null},{"position":22,"raw":"[Zeiler and Fergus, 2014] Matthew D Zeiler and Rob Fergus. Visualizing and understanding convolutional networks. In ECCV, pages 818-833. Springer, 2014.","title":"Visualizing and understanding convolutional networks","authors":"D Matthew, Rob Zeiler, Fergus","year":"2014","venue":"ECCV","doi":null,"url":null,"internal_url":null},{"position":23,"raw":"[Zhang et al., 2017] Pingping Zhang, Dong Wang, Huchuan Lu, Hongyu Wang, and Xiang Ruan. Amulet: Aggregating multi-level convolutional features for salient object detec- tion. In ICCV, Oct 2017.","title":"Amulet: Aggregating multi-level convolutional features for salient object detection","authors":"Pingping Zhang, Dong Wang, Huchuan Lu, Hongyu Wang, Xiang Ruan","year":"2017","venue":"ICCV","doi":null,"url":null,"internal_url":null}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}