Human Action Recognition in Table-Top Scenarios : An HMM-Based Analysis to Optimize the Performance
Hidden Markov models have been extensively and successfully used for the recognition of human actions. Though there exist well-established algorithms to optimize the transition and output probabilities, the type of features to use and specifically the number of states and Gaussian have to be chosen manually. Here we present a quantitative study on selecting the optimal feature set for recognition of simple object manipulation actions pointing, rotating and grasping in a table-top scenario. This study has resulted in recognition rate higher than 90%. Also three different parameters, namely the number of states and Gaussian for HMM and the number of training iterations, are considered for optimization of the recognition rate with 5 different feature sets on our motion capture data set from 10 persons.
KeywordsHidden Markov model Action Recognition Optimization
Unable to display preview. Download preview PDF.
- 2.Ascension. Motion Star Real-Time Motion Capture. http://www.ascension-tech.com/products/motionstar_10_04.pdf
- 4.Calinon, S., Billard, A., Guenter, F.: Discriminative and adaptative imitation in uni-manual and bi-manual tasks. Robotics and Autonomous Systems 54 (2005)Google Scholar
- 5.Grest, D., Koch, R., Krueger, V.: Single view motion tracking by depth and silhoutte information. In: Scandinavian Conference on Image Analysis (2007)Google Scholar
- 6.Guenter, S., Bunke, H.: Optimizing the number of states and training iterations and gaussians in an hmm-based handwritten word recognizer. In: Seventh International Conference on Document Analysis and Recognition, vol. 1, pp. 472–476 (August 2003)Google Scholar
- 7.Mataric, M.J.: Sensory-motor primitives as a basis for imitation: linking perception to action and biology to robotics. In: Dautenhahn, K., Nehaniv, C.L. (eds.) Imitation in Animals and Artifacts, pp. 391–422. MIT Press, Cambridge (2002)Google Scholar
- 9.Krueger, V., Grest, D.: Using hidden markov models for recognizing action primitives in complex actions. In: Scandinavian Conference on Image Analysis (2007)Google Scholar
- 10.Campbell, L., Bobick, A.: Recognition of human body motion using phase space constraints. In: International Conference in Computer Vision, pp. 624–630 (1995)Google Scholar
- 12.Murphy, K.: Hidden Markov Model (HMM) Toolbox for Matlab (1998), http://www.cs.ubc.ca/~murphyk/Software/HMM/hmm.html
- 13.Newtson, D., Engquist, G., Bois, J.: The objective basis of behavior unit. Journal of Personality and social psychology, 847–862 (1977)Google Scholar
- 15.Vicente, I.S., Kragic, D.: Learning and recognition of object manipulation actions using linear and nonlinear dimensionality reduction. In: 15th IEEE Int. Symp. on Robot and Human Interactive Communication (RO-MAN) (submitted, 2007)Google Scholar