Application of Transfer Learning in RGB-D Object Recognition

被引:0
作者
Kumar, Abhishek [1 ]
Shrivatsav, S. Nithin [2 ]
Subrahmanyam, G. R. K. S. [3 ]
Mishra, Deepak [3 ]
机构
[1] Vellore Inst Technol, Sch Elect Engn, Vellore 632014, Tamil Nadu, India
[2] Natl Inst Technol, Dept EEE, Tiruchirappalli 620015, Tamil Nadu, India
[3] Dept Avion IIST, Thiruvananthapuram 695547, Kerala, India
来源
2016 INTERNATIONAL CONFERENCE ON ADVANCES IN COMPUTING, COMMUNICATIONS AND INFORMATICS (ICACCI) | 2016年
关键词
Computer Vision; Deep Learning; Convolution Neural Network; Transfer Learning;
D O I
暂无
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
In this work, we apply Transfer Learning for a Multimodal Deep learning network for fast and robust object recognition using RGB-D dataset. The ability for a network to train quickly and recognize objects robustly is very important in the field of Robotics. The Multimodal deep learning network avoids time-consuming hand-crafted features and makes use of a RGB-D architecture for robust object recognition. Our architecture has two important features. First, it makes use of both RGB and Depth information of an image to recognize it. To achieve this, our architecture has two CNN processing streams, one for RGB modality and the other for the depth modality. This enables the network to achieve higher accuracy than normal single stream RGB network. We encoded the depth image into colour image before passing it into the CNN stream. The other important feature is the speed of training and improving the accuracy further. To achieve this, we made use of Transfer learning. Firstly we trained a CNN network with 10 classes of different objects and then we transfer the parameters to RGB and depth CNN network. This enables the network to train faster and also achieve higher accuracy for a given number of epochs.
引用
收藏
页码:580 / 584
页数:5
相关论文
共 17 条
  • [11] Freeman WT, 1999, COMPUT GRAPHICS-US, V33, P65, DOI 10.1145/345370.345417
  • [12] ImageNet Classification with Deep Convolutional Neural Networks
    Krizhevsky, Alex
    Sutskever, Ilya
    Hinton, Geoffrey E.
    [J]. COMMUNICATIONS OF THE ACM, 2017, 60 (06) : 84 - 90
  • [13] Gradient-based learning applied to document recognition
    Lecun, Y
    Bottou, L
    Bengio, Y
    Haffner, P
    [J]. PROCEEDINGS OF THE IEEE, 1998, 86 (11) : 2278 - 2324
  • [14] Liu T., 2015, Implementation of Training Convolutional Neural Networks
  • [15] Pan SinnoJialin., 2012, Transfer learning with applications
  • [16] Srivastava N, 2014, J MACH LEARN RES, V15, P1929
  • [17] A survey of transfer learning
    Weiss K.
    Khoshgoftaar T.M.
    Wang D.D.
    [J]. Journal of Big Data, 3 (1)