GAN-Based Data Augmentation for Visual Finger Spelling Recognition

被引:2
作者
Kwolek, Bogdan [1 ]
机构
[1] AGH Univ Sci & Technol, 30 Mickiewicza Av, PL-30059 Krakow, Poland
来源
ELEVENTH INTERNATIONAL CONFERENCE ON MACHINE VISION (ICMV 2018) | 2019年 / 11041卷
关键词
Generative Adversarial Networks; CNNs; Finger Spelling Recognition; HAND GESTURE RECOGNITION; REAL-TIME;
D O I
10.1117/12.2522935
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
In this work we extend WGAN-GP in order to achieve better generation of synthesized images for finger spelling classification. The main difference between the ordinary WGAN-GP and the proposed algorithm is that in the training we employ both training samples and training labels. These training labels are fed to the generator, that generates the synthetic images using both the randomized latent input and the input label. In ordinary WGAN-GP, latent input variables are usually sampled from an unconditional prior. In the proposed algorithm the latent input vector is a concatenation of random part, the class labels and additional variables that are drawn from Gaussian distributions representing hand poses or gesture attributes. The JSL dataset for Hiragana sign recognition has been balanced using the rendered samples on the basis of a 3D hand model as well as the extended WGAN-GP.
引用
收藏
页数:8
相关论文
共 25 条
[1]  
[Anonymous], TECH REP
[2]  
[Anonymous], MULTICHANNEL CONVOLU
[3]  
[Anonymous], ADV NEURAL INFORM PR
[4]  
[Anonymous], 2016, NEURAL COMPUT APPL
[5]  
[Anonymous], 2016, ABS160706450 CORR
[6]  
Arjovsky M, 2017, PR MACH LEARN RES, V70
[7]   Hand Gesture Modeling and Recognition for Human and Robot Interactive Assembly Using Hidden Markov Models [J].
Chen, Fei ;
Zhong, Qiubo ;
Cannella, Ferdinando ;
Sekiyama, Kosuke ;
Fukuda, Toshio .
INTERNATIONAL JOURNAL OF ADVANCED ROBOTIC SYSTEMS, 2015, 12
[8]  
Cheung V., 2016, P NIPS, P2234
[9]  
Goodfellow IJ, 2014, ADV NEUR IN, V27, P2672, DOI DOI 10.1145/3422622
[10]  
Gulrajani F., 2017, Improved training of wasserstein gans, P5767