A Recurrent Temporal Model for Semantic Levels Categorization Based on Human Visual System

Figure 3

The accuracy and reaction time for the three levels of superordinate, basic, and the subordinate for short and long ISI. (a) The accuracy for three levels of categorization. It can be seen that the superordinate is more accurate than other levels for both long and short ISI. (b) Unlike accuracy, reaction times for superordinate level are less than other levels, which means the superordinate is the fastest level. These two images show the superordinate advantage for presence and absence of top-down influences.

(a)

(b)