A natural way to represent a spatiotemporal events is to consider input image sequence as data in space and time [13, 23] by associating the serial order of the pattern with the dimensionality of the pattern vector. The first temporal event is represented in the plane t=0 and the second temporal event by plane t=1, and so on. The entire spatiotemporal pattern vector is considered as a whole by the framework. Figure 2 shows an example in which the hand sign ``no'' is represented by a spatiotemporal sequence (three images).
Figure 2:
The sign ``no'' and its image sequence representation. (a)
The sign of ``no'', snap middle finger, index, and thumb together. (b) The
sequence representation of the sign ``no''.