Fig. 2: Schematic diagram of the continuous-time entropy-limit minimization (CTELM) and constrained-CTELM (CCTELM) models. | Communications Biology

Fig. 2: Schematic diagram of the continuous-time entropy-limit minimization (CTELM) and constrained-CTELM (CCTELM) models.

From: Human visual search follows a suboptimal Bayesian strategy revealed by a spatiotemporal computational model and experiment

Fig. 2: Schematic diagram of the continuous-time entropy-limit minimization (CTELM) and constrained-CTELM (CCTELM) models.

The model starts each trial by fixating at the image center. Then, at each time step, it accumulates the visual information sampled since the start of the current fixation, and calculates the posterior probability map of target location across the search field by Bayesian theory. If the posterior probability of target being at current fixation location is larger than a target detection threshold, the model will set the target location as the current fixation location and terminate search. Otherwise, if the posterior probability at current fixation location is lower than the saccade threshold, it will saccade to a new location. If neither of the thresholds are met, the model will continue the current fixation. The whole process iterates until the model finds the target.

Back to article page