Large Scale Novel Object Discovery in 3D

被引:3
作者
Srivastava, Siddharth [1 ]
Sharma, Gaurav [2 ]
Lall, Brejesh [1 ]
机构
[1] Indian Inst Technol, Delhi, India
[2] Indian Inst Technol, Kanpur, Uttar Pradesh, India
来源
2018 IEEE WINTER CONFERENCE ON APPLICATIONS OF COMPUTER VISION (WACV 2018) | 2018年
关键词
RECOGNITION; SURFACE;
D O I
10.1109/WACV.2018.00026
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
We present a method for discovering never-seen-before objects in 3D point clouds obtained from sensors like Microsoft Kinect. We generate supervoxels directly from the point cloud data and use them with a Siamese network, built on a recently proposed 3D convolutional neural network architecture. We use known objects to train a non-linear embedding of supervoxels, by optimizing the criteria that supervoxels which fall on the same object should be closer than those which fall on different objects, in the embedding space. We test on unknown objects, which were not seen during training, and perform clustering in the learned embedding space of supervoxels to effectively perform novel object discovery. We validate the method with extensive experiments, quantitatively showing that it can discover numerous unseen objects while being trained on only a few dense 3D models. We also show very good qualitative results of object discovery in point cloud data when the test objects, either specific instances or even categories, were never seen during training.
引用
收藏
页码:179 / 188
页数:10
相关论文
共 38 条
[21]   Indoor Scene Understanding with RGB-D Images: Bottom-up Segmentation, Object Detection and Semantic Segmentation [J].
Gupta, Saurabh ;
Arbelaez, Pablo ;
Girshick, Ross ;
Malik, Jitendra .
INTERNATIONAL JOURNAL OF COMPUTER VISION, 2015, 112 (02) :133-149
[22]   Learning Rich Features from RGB-D Images for Object Detection and Segmentation [J].
Gupta, Saurabh ;
Girshick, Ross ;
Arbelaez, Pablo ;
Malik, Jitendra .
COMPUTER VISION - ECCV 2014, PT VII, 2014, 8695 :345-360
[23]   Perceptual Organization and Recognition of Indoor Scenes from RGB-D Images [J].
Gupta, Saurabh ;
Arbelaez, Pablo ;
Malik, Jitendra .
2013 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2013, :564-571
[24]  
Herbst E, 2011, IEEE INT CONF ROBOT, P2623
[25]  
Jiwon Shin, 2010, 2010 IEEE International Conference on Robotics and Automation (ICRA 2010), P5041, DOI 10.1109/ROBOT.2010.5509914
[26]   Using spin images for efficient object recognition in cluttered 3D scenes [J].
Johnson, AE ;
Hebert, M .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1999, 21 (05) :433-449
[27]  
Karpathy A, 2013, IEEE INT CONF ROBOT, P2088, DOI 10.1109/ICRA.2013.6630857
[28]  
Maturana D, 2015, IEEE INT C INT ROBOT, P922, DOI 10.1109/IROS.2015.7353481
[29]  
Mueller CA, 2016, IEEE INT CONF ROBOT, P2263, DOI 10.1109/ICRA.2016.7487375
[30]   Voxel Cloud Connectivity Segmentation - Supervoxels for Point Clouds [J].
Papon, Jeremie ;
Abramov, Alexey ;
Schoeler, Markus ;
Woergoetter, Florentin .
2013 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2013, :2027-2034