Although Convolutional Neural Networks performed better in object detection, CNNs does not care about spatial relationships existing in an image. In this paper, we try describe "capsule network based object detection" model COD based on the VGG16 model (as a base network), which presents a substantial result in many sections of object detection over Convolution Neural Network based model by achieving the problem of spatial relationships. We used matrix capsules and dynamic EM routing to classify object from different viewpoints. The whole model is grounded on "dynamic routing between capsules", which is suggested by Geoffrey E Hinton. Both proposed theories use capsules that maps feature properties of an object as information for detecting that object which is extracted by capsules and Dynamic routing groups the capsules of lower level into parent level capsules by an iterative dynamic routing process. We train and test our model on Pascal VOC 2007 and dataset. We implement this in python using Keras (Tensorflow as backend) and train our model in Google cloud compute engine. COD achieves an accuracy of 67.3 mAP on Pascal VOC-2007 dataset and performing a comparable performance with Fast R-CNN.

Object Detection, CNN, Capsule Networks, VGG16

[1] S. Ren, K. He, R. Girshick and J. Sun, "Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks," IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017.
[2] M. Everingham, S. M. A. Eslami, L. Van Gool, C. K. I. Williams, J. Winn and A. Zisserman, "The Pascal Visual Object Classes Challenge: A Retrospective," International Journal of Computer Vision, 2014.
[3] Jia Deng, Wei Dong, R. Socher, Li-Jia Li, Kai Li and Li Fei-Fei, "ImageNet: A large-scale hierarchical image database," IEEE Conference on Computer Vision and Pattern Recognition, 2009.
[4] J. Dai, Y. Li, K. He and J. Sun, "R-FCN: Object Detection via Region-based Fully Convolutional Networks," 2016.
[5] W. Liu, D. Anguelov, D. Erhan, C. Szegedy, S. Reed, C.-Y. Fu and A. C. Berg, "SSD: Single Shot MultiBox Detector" IEEE European Conference on Computer Vision, 2015.
[6] J. Redmon, S. Divvala, R. Girshick and A. Farhadi, "You Only Look Once: Unified, Real-Time Object Detection," 2015.
[7] Ross Girshick, "Fast RCNN" IEEE International Conference on Computer Vision, 2015.
[8] R. Girshick, J. Donahue, T. Darrell and J. Malik, "Rich feature hierarchies for accurate object detection and semantic segmentation," IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2014.
[9] S. Sabour, N. Frosst and G. E. Hinton, "Dynamic Routing Between Capsules," 2017.
[10] G. E. Hinton, A. Krizhevsky and S. D. Wang, "Transforming Auto-encoders," in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2011.
[11] D. Wang and Q. Liu, “An Optimization View on Dynamic Routing between Capsules" Workshop track in International Conference on Learning Representations 2018.
[12] E. Xi, S. Bing and Y. Jin, "Capsule Network Performance on Complex Data," 2017.
[13] A. Jaiswal, W. AbdAlmageed and P. Natarajan, "CapsuleGAN: Generative Adversarial Capsule Network," 2018.
[14] A. G. Howard, M. Zhu, B. Chen, D. Kalenichenko, W. Wang, T. Weyand, M. Andreetto and H. Adam, "MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications" 2017.
[15] J. R. R. Uijlings, K. E. A. Van De Sande, T. Gevers and A. W. M. Smeulders, "Selective Search for Object Recognition" International Journal of Computer Vision 2013.
[16] J. Long, E. Shelhamer and T. Darrell, "Fully convolutional networks for semantic segmentation," IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2015.
[17] K. Simonyan and A. Zisserman, "Very Deep Convolutional Networks for Large-Scale Image Recognition," 2015.
[18] L. Zhu and H. Yuan, "Spatial Relationship for Object Recognition" International Conference on Learning Representations 2015.
[19] Hoiem, D., Chodpathumwan, Y., Dai, “Diagnosing error in object detectors”, IEEE European Conference on Computer Vision, 2012.