Conditional Variational Capsule Network for Open Set Recognition

被引:34
作者
Guo, Yunrui [1 ,2 ]
Camporese, Guglielmo [2 ]
Yang, Wenjing [1 ]
Sperduti, Alessandro [2 ]
Ballan, Lamberto [2 ]
机构
[1] Natl Univ Def Technol, Changsha, Peoples R China
[2] Univ Padua, Dept Math Tullio Levi Civita, Padua, Italy
来源
2021 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2021) | 2021年
关键词
D O I
10.1109/ICCV48922.2021.00017
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
In open set recognition, a classifier has to detect unknown classes that are not known at training time. In order to recognize new categories, the classifier has to project the input samples of known classes in very compact and separated regions of the features space for discriminating samples of unknown classes. Recently proposed Capsule Networks have shown to outperform alternatives in many fields, particularly in image recognition, however they have not been fully applied yet to open-set recognition. In capsule networks, scalar neurons are replaced by capsule vectors or matrices, whose entries represent different properties of objects. In our proposal, during training, capsules features of the same known class are encouraged to match a pre-defined gaussian, one for each class. To this end, we use the variational autoencoder framework, with a set of gaussian priors as the approximation for the posterior distribution. In this way, we are able to control the compactness of the features of the same class around the center of the gaussians, thus controlling the ability of the classifier in detecting samples from unknown classes. We conducted several experiments and ablation of our model, obtaining state of the art results on different datasets in the open set recognition and unknown detection tasks.
引用
收藏
页码:103 / 111
页数:9
相关论文
共 34 条
[11]  
Hongjie Zhang, 2020, Computer Vision - ECCV 2020. 16th European Conference. Proceedings. Lecture Notes in Computer Science (LNCS 12348), P102, DOI 10.1007/978-3-030-58580-8_7
[12]  
Jain LP, 2014, LECT NOTES COMPUT SC, V8691, P393, DOI 10.1007/978-3-319-10578-9_26
[13]  
Kingma D.P, TRACK 2014 PROC 2 IN
[14]  
Krizhevsky A., 2009, LEARNING MULTIPLE LA
[15]   ImageNet Classification with Deep Convolutional Neural Networks [J].
Krizhevsky, Alex ;
Sutskever, Ilya ;
Hinton, Geoffrey E. .
COMMUNICATIONS OF THE ACM, 2017, 60 (06) :84-90
[16]  
LeCun Y, 2010, THE MNIST DATABASE of handwritten digits, DOI DOI 10.1007/S11063-009-9095-3
[17]   Nearest neighbors distance ratio open-set classifier [J].
Mendes Junior, Pedro R. ;
de Souza, Roberto M. ;
Werneck, Rafael de O. ;
Stein, Bernardo V. ;
Pazinato, Daniel V. ;
de Almeida, Waldir R. ;
Penatti, Otavio A. B. ;
Torres, Ricardo da S. ;
Rocha, Anderson .
MACHINE LEARNING, 2017, 106 (03) :359-386
[18]  
Neal L., 2018, PROC EUROPEAN C COMP
[19]  
Netzer Y, 2011, READING DIGITS NATUR
[20]  
Nguyen HH, 2019, INT CONF ACOUST SPEE, P2307, DOI 10.1109/ICASSP.2019.8682602