Purchase this article with an account.
Ipek Oruc, Jason Barton; Brief adaptation increases sensitivity of face recognition. Journal of Vision 2008;8(6):1143. doi: https://doi.org/10.1167/8.6.1143.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Some models of adaptation propose that adaptation may serve to enhance discriminative sensitivity in the visual system. Increased sensitivity has been well established for retinal adaptation. However, empirical data for higher-levels remain inconclusive, with some studies appearing to confirm enhancement for the discrimination of orientation, contrast, and direction of motion, and others failing to find such improvements. In this study we assessed the impact of variable periods of adaptation on face recognition. We measured discrimination contrast thresholds for faces in a five-alternative forced-choice paradigm with or without prior adaptation to a face, for periods ranging from 10ms to 6400ms. Following adaptation and a noise mask lasting 50ms, subjects saw a low-contrast test face lasting 150ms. Contrast of the test image was varied with the Quest procedure to estimate 82% thresholds. For adaptation periods greater than 500ms, contrast thresholds for both faces identical to and different from the adapting face were elevated and continued to increase with longer adaptation durations. However, for durations less than 500ms, adaptation reduced thresholds for the same face, indicating a facilitation effect that was maximal at 200ms, but continued to elevate thresholds for different faces. Similar effects were obtained with adapting faces that differed in size from the test faces, excluding adaptation to low-level image properties as the source of the results. To exclude a response-bias account of the facilitation result we repeated the same experiment with a two-interval two-alternative detection task using the same stimulus set, with similar findings at short adapting durations. We conclude that brief periods of adaptation may serve to enhance recognition in high-level object processing.
This PDF is available to Subscribers Only