Purchase this article with an account.
Irving Biederman, Tianyi Zhu, Miles Nelken, Emily X Meschke, Catrina M Hacker; The Cost of Matching Depth-Rotated Faces: A Simple Function of Image Similarity. Journal of Vision 2019;19(10):136b. doi: https://doi.org/10.1167/19.10.136b.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
When unfamiliar faces are to be recognized or matched at different orientations in depth, sizeable costs in the speed and accuracy of performance have been documented attributable to the disparity in orientations. There have been no rigorous, quantitative assessments of these costs. Here we evaluated the effects of orientation disparity in a minimal match-to-sample paradigm of a triangular display of three faces, with one of the two test faces physically identical to the sample (Fig.1, left). The other test face was a foil. Both test faces had the same orientation in depth at an angle that differed by 0° to 20° from the sample (Fig. 1, right). The dissimilarities of the faces were scaled by the Gabor jet model (Lades et al., 1993), a model based on V1 simple cell tuning that predicts psychophysical similarity almost perfectly when the faces are at the same orientation (Yue et al., 2012). The greater the dissimilarity of the rotated matching face to the sample and the greater the similarity of the foil and matching test faces the longer the reaction times (RTs). The second of these two factors—the similarity of matching to foil test faces—had three times the effect per unit of Gabor similarity than the dissimilarity of the matching test face to the sample. These two costs were additive and were sufficient to account for all the costs of orientation disparity (Fig. 2). The design also allowed a test of whether individuals who incurred greater costs on RTs of increasing similarity of foil-to-matching faces at 0° would also suffer greater costs when the similarity of the matching to sample face was reduced because of rotation disparity. They did not, suggesting that the capacity for discrimination of highly similar faces is independent of the capacity for achieving invariance over viewpoint.
This PDF is available to Subscribers Only