Abstract
In the field of computer vision, there are two emerging approaches that have drawn much attention, and they have recently become popular way to solve various kinds of recognition problem. The first approach is unsupervised feature learning based on deep learning technique, and second approach is to conduct recognition using depth information thank to recent progress in depth sensor. At this point, it seems reasonable that one is curious about effectiveness of deep learning from raw depth data. However, a few researches have attempted to learn depth features with a deep network, and the validity has not been well studied in terms of quantitative analysis. To this end, we learned depth features for human activity recognition using existing deep learning algorithm and evaluated effectiveness of the learned depth feature on activity recognition. Furthermore, we provide analysis in detail and valuable discussion with additional experiments.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Hinton, G., Osindero, S., Teh, Y.: A fast learning algorithms for deep belief nets. Neu. Comp. (2006)
Salakhutdinov, R., Hinton, G.: Deep Boltzmann Machines. In: International Conference on AI and Statistics (2009)
Lee, H., Grosse, R., Ranganath, R., Ng, A.: Convolutional Deep Belief Networks for Scalable Unsupervised Learning of Hierarchical Representations. In: ICML (2009)
Bengio, Y., Lamblin, P., Popovici, D., Larochelle, H.: Greedy layerwise training of deep networks. In: NIPS (2006)
Lowe, D.: Distinctive image features from scale-invariant keypoints. IJCV (2004)
Dalal, N., Triggs, B.: Histograms of oriented gradients for human detection. In: CVPR (2005)
Hinton, G., Osindero, S., Teh, Y.: A Fast Learning Algorithm for Deep Belief Nets. Neural Computation 18(7), 1527–1554 (2006)
Bo, L., Ren, X., Fox, D.: Hierarchical Matching Pursuit for Image Classification: Architecture and Fast Algorithms. In: NIPS (2011)
Yu, K., Lin, Y., Lafferty, J.: Learning Image Representations from the Pixel Level via Hierarchical Sparse Coding. In: CVPR (2011)
Le, Q.V., Zou, W.Y., Yeung, S.Y., Ng, A.Y.: Learning hierarchical spatio-temporal features for action recognition with independent subspace analysis. In: CVPR (2011)
Shotton, J., Fitzgibbon, A., Cook, M., Sharp, T., Finocchio, M., Moore, R., Kipman, A., Blake, A.: Real-time human pose recognition in parts from single depth images. In: IEEE CVPR (2011)
Koppula, H.S., Saxena, A.: Learning Spatio-Temporal Structure from RGB-D Videos for Human Activity Detection and Anticipation. In: ICML (2013)
Sung, J., Ponce, C., Selman, B., Saxena, A.: Unstructured human activity detection from rgbd images. In: ICRA (2012)
Socher, R., Huval, B., Bath, B.P., Manning, C.D., Ng, A.Y.: Convolutional-Recursive Deep Learning for 3D Object Classification. In: NIPS (2012)
Wang, H., Ullah, M.M., Klaser, A., Laptev, I., Schmid, C.: Evaluation of local spatio-temporal features for action recognition. In: BMVC (2010)
Hyvarinen, A., Hurri, J., Hoyer, P.: Natural Image Statistics. Springer (2009)
Pierre, C.: Independent Component Analysis: a new concept? Signal Processing 36(3), 287–314 (1994)
Wu, T.F., Lin, C.J., Weng, R.C.: Probability estimates for multi-class classification by pairwise coupling. Journal of Machine Learning Research 5, 975–1005 (2003)
http://www.codeproject.com/Articles/317974/KinectDepthSmoothing
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2013 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Jang, J., Park, Y., Suh, I.H. (2013). Empirical Evaluation on Deep Learning of Depth Feature for Human Activity Recognition. In: Lee, M., Hirose, A., Hou, ZG., Kil, R.M. (eds) Neural Information Processing. ICONIP 2013. Lecture Notes in Computer Science, vol 8228. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-42051-1_71
Download citation
DOI: https://doi.org/10.1007/978-3-642-42051-1_71
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-42050-4
Online ISBN: 978-3-642-42051-1
eBook Packages: Computer ScienceComputer Science (R0)