Purchase this article with an account.
D. Yu, S. T. L. Chung; Critical Orientation for Face Identification in People with Central Vision Loss. Invest. Ophthalmol. Vis. Sci. 2010;51(13):5170.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Difficulty in identifying faces is a common complaint of people with central vision loss. Dakin and Watt (2009) reported that the horizontal component of face images is most informative for face identification in young adults with normal vision. In this study, we examined whether or not people with central vision loss similarly rely primarily on the horizontal component of face images for face identification.
Five observers with central vision loss (aged 57 to 82, acuity: 0.4 to 1.1 logMAR) and three age-matched observers with normal vision (aged 58 to 73) participated in the study. They first previewed a set of face images of 294 celebrities and gave a familiarity rating for each image. Only faces that were rated as "very familiar" were used subsequently. For each of the "very familiar" faces, a different face image of the same celebrity was used as the test face. We applied an orientation filter (bandwidth=23°) to restrict information contained in the test faces. The center of the filter ranged from 0° (horizontal) to 150° in steps of 30°. Identification accuracy for these filtered faces was measured (the unfiltered condition was also tested for comparison). Between 8 and 20 trials per orientation band were tested for each observer.
Consistent with Dakin and Watt (2009), the accuracy for identifying filtered face images of the age-matched observers was highest (averaged 76%, compared with 87% for the unfiltered condition) at 0° filter orientation (horizontal), dropping monotonically as the filter orientation deviated systematically from horizontal, and was the lowest (averaged 32%) at 90° filter orientation (vertical). In comparison, all observers with central vision loss showed (1) a flatter accuracy vs. filter orientation function, (2) a larger difference in accuracy between peak performance and the unfiltered condition, and (3) a higher accuracy for images containing only the vertical information than those containing some oblique orientations.
The reliance on information along different orientations in face images for face identification is different between people with normal vision and those with central vision loss. While the horizontal information alone can support reasonably good performance for identifying faces in people with normal vision, information along other orientations seems to be similarly informative for people with central vision loss.
This PDF is available to Subscribers Only