×
This animation demonstrates a framework for retrieving and visualizing biologically and psychologically relevant information form biological motion patterns. It is based on walking data from 50 male and 50 female walkers. Using a motion capture system their movement were recorded while walking on a treadmill.
The data were subsequently transformed into a representation which allows for linear morphing. The resulting "walking space" was then transformed usig principal component analysis. A space spanned by the first 10 eigenwalkers was used to compute linear discriminant functions for the respective attributes.
Sex and weight of each walker were directly available from our records. The outer two attributes were derived from psychological experiments. A number of observers were presented with point-light displays of the 100 walkers. For each of them they had to rate the attributes nervous/relaxed and happy/sad on a scale of 6 steps.
The procedure is described in dertail in these papers:
Troje, N. F. (2002), Decomposing biological motion: A framework for analysis and synthesis of human gair patterns, Journal of Vision, 2:371-387.
Troje, N. F. (2002) The little difference: Fourier based gender classification from biological motion; In: Dynamic Perception, R. P. Würtz and M. Lappe (eds), Aka Press, Berlin:115-120
Troje, N. F. (2008) Retrieving information from human movement patterns. In: Shipley, T. F. and Zacks, J. M. (eds.) Understanding Events: How Humans See, Represent, and Act on Events. Oxford University Press, pp. 308-334.
