Combining Two Visual Cognition Systems Using Confidence Radius and Combinatorial Fusion

When to Combine Two Visual Cognition Systems

“…The object is constructed of metal washers connected together in an irregular manner to minimize movement upon reaching the ground, while being large enough to be visible while airborne. This landscape and test environment is modeled on the one used in [2], [3], [13], [14], [17], but with enhanced visual clarity for the participant. The rectangular grid is marked by yellow flags so that observers can clearly see the boundary of the region.…”

“…More recently, Koriat [9] stipulated that choosing the more confident participant in a pair, without information exchange between the two individuals, could outperform either participant. While Koriat analyzes joint decisions between non-communicating participants, Batallones et al [2], [3], McMunn-Coffran [13], [14], and Paolercio et al [11] expand the works of Bahrami and Ernst by further optimizing joint visual cognition decisions.…”

“…Subjects observe a small object being thrown in a fashion similar to previous experiments [2], [3], [13], [14], [17], where they are separately asked of the object's perceived landing point. These subjects are treated as two visual cognition systems in the physical space, and are connected to two scoring systems in the common visual space.…”

Joint decision-making on two visual perception systems

“…The common visual space we provide is similar to the concept of dialectical bootstrapping proposed by Hertwig and Hertzog, in that to improve the accuracy of an estimation, the estimates must bracket a true value, represented in our experiment by the known landing site of the token. We extend the visual space and increase the number of intervals to reduce systematic error and random error (fluctuations in volunteers' performances) [4] Combinatorial fusion has been shown to be more accurate in many cases when compared to the mean and weighted means [2], [3], [11], [12], [13]. We apply combinatorial fusion to the process of combining two visual cognition systems in cases where it is unknown whether one or the other has the better perception of the visual field.…”

“…The actual landing site provides a means of measuring how accurate the guesses are, so that combinatorial fusion can be applied to cases where the truth remains an unknown. Work in this area has been done by one of us and his colleagues [2], [3], [11], [12], [13]. Our current work takes into consideration cognition diversity between and relative performance among these two visual cognition systems.…”