人類骨架辨識在電腦視覺上是一個新興且熱門的研究，有很多的應用像是自動監控系統、人機互動等等。由於人體相對於一般的物體偵測方法更具挑戰性，為了能在複雜的動作型態下，得到正確的人體節點位置，我們繼承Convol-utional Pose Machine的架構並利用卷積神經網路強健的分類能力，有效且準確的學習不同人體節點位置的特徵，找出節點座標和特徵之間的關聯性，並利用此關聯性來預測人體骨架位置。為了增加辨識速度，我們先做背景去除，找出人體大致的輪廓與中心點，再輸入卷積類神經網路進行辨識，實驗結果顯示我們所提出的系統能有效提高30%辨識的速度。

The human pose estimation is an emerging research topic in the area of computer vision, which has many particular applications including automated surveillance systems, interactive human machine interfaces, and video-based pose recognitions. The human pose estimation is an advanced localization technique, of which implementation is beyond the object detection because of the possibility of articulated human body and various occlusions. We inherent the structure the Convol-utional Pose Machine.In order to get the accurate localization of the human joints, this study implements a convolution neural network to model the features in human pose training images by operating different-sized kernels to acquire the optimum features. Then, the trained model is utilized to localize the human joints within the test images after segmenting the foreground through the ViBe technique.

[1] L. Xia, C.-C. Chen, and J. K. Aggarwal,“ Human detection using depth information by Kinect,” Computer Vision and Pattern Recognition Workshops, pp.15-22, June, 2011.
[2] Pedro F. Felzenszwalb, Daniel P. Huttenlocher,“Pictorial structures for object recognition,” International Journal of Computer Vision, Volume 61,Issue 1, pp.55-79, January, 2005.
[3] Jia, Yangqing and Shelhamer, Evan and Donahue, Jeff and Karayev, Sergey and Long, Jonathan and Girshick, Ross and Guadarrama, Sergio and Darrell, Trevor, “Caffe: Convolutional Architecture for Fast Feature Embedding,” The 22nd ACM international conference on Multimedia, pp.675-678, November, 2014.
[4] Alexander Toshev and Christian Szegedy, “Deep pose: Human pose estimation via deep neural networks,” IEEE Conference on Computer Vision and Pattern Recognition, pp.23-28, June, 2014.
[5] Xiaochuan Fan, Kang Zheng, Yuewei Lin, Song Wang, “Combining Local Appearance and Holistic View: Dual-Source Deep Neural Networks for Human Pose Estimation,Computer ,” IEEE Conference on Computer Vision and Pattern Recognition, pp.7-12, June, 2015.
[6] Shih-En Wei, Varun Ramakrishna, Takeo Kanade, Yaser Sheikh “Convolutional Pose Machines,” IEEE Conference on Computer Vision and Pattern Recognition, July, 2016.
[7] Varun Ramakrishna Daniel Munoz Martial Hebert J. Andrew (Drew) Bagnell Yaser Sheikh “Pose Machines: Articulated Pose Estimation via Inference Machines,” European Conference on Computer Vision, pp.33-47, September, 2014.
[8] Olivier Barnich and Marc Van Droogenbroeck, “ViBe: A universal background subtraction algorithm for video sequences,” IEEE Transactions on Image Processing ,Volume20,Issue 6, pp. 1709 - 1724 , June, 2011.
[9] Yi Yang, Deva Ramanan, “Articulated Pose Estimation with Flexible Mixtures of Parts,” IEEE Conference on Computer Vision and Pattern Recognition, pp.20-25, June, 2011.
[10] Stanford CS class note: Convolutional Neural Networks for Visual Recognition.
[Online]. Available: http://cs231n.stanford.edu/
[11] Convolutional neural network.
[Online]. Available: https://en.wikipedia.org/wiki/Convolutional_neural_network
[12] Y. LeCun, L. Bottou, Y. Bengio, and P. Haffner, “Gradient-based learning applied to document recognition,” Proceedings of the IEEE, Volume 86, Issue 11, pp.2278-2324, August, 2002.
[13] Brian Cheung, Carl Sable, “Hybrid Evolution of Convolutional Networks,” Machine Learning and Applications and Workshops, Volume 1,pp.293-297, December, 2011.
[14] Mykhaylo Andriluka and Leonid Pishchulin and Peter Gehler and Schiele, Bernt, “2D Human Pose Estimation: New Benchmark and State of the Art Analysis” IEEE Conference on Computer Vision and Pattern Recognition, pp.3686 – 3693, June, 2014.
[15] Johnson, Sam and Everingham, Mark, “Clustered Pose and Nonlinear Appearance Models for Human Pose Estimation,” Proceedings of the British Machine Vision Conference,2010
[16] Girshick, Ross and Donahue, Jeff and Darrell, Trevor and Malik, Jitendra,“Rich feature hierarchies for accurate object detection and semantic segmentation,” IEEE Conference on Computer Vision and Pattern Recognition, pp.23-28, June 2014.