About Face: Matching Unfamiliar Faces Across Rotations of View and Lighting

Abstract

Matching the identities of unfamiliar faces is heavily influenced by variations in their images. Changes to viewpoint and lighting direction during face perception are commonplace across yaw and pitch axes and can result in dramatic image differences. We report two experiments that, for the first time, factorially investigate the combined effects of lighting and view angle on matching performance for unfamiliar faces. The use of three-dimensional head models allowed control of both lighting and viewpoint. We found viewpoint effects in the yaw axis with little to no effect of lighting. However, for rotations about the pitch axis, there were both viewpoint and lighting effects and these interacted where lighting effects were found only for front views and views from below. The pattern of effects was similar regardless of whether view variation occurred as a result of head (Experiment 1) or camera (Experiment 2) suggesting that face matching is not purely image based. Along with face inversion effects in Experiment 1, the results of this study suggest that face perception is based on shape and surface information and draws on implicit knowledge of upright faces and ecological (top) lighting conditions.

Keywords: face perception; lighting; unfamiliar face matching; viewpoint.

Figure 1.

Examples of stimuli in which the head was rotated about the yaw axis (left column 45° leftwards, middle column 0° and right column 45° rightwards) and lighting rotated about the yaw axis (top row 45° leftwards, middle row 0° and bottom row 45° rightwards).

Figure 2.

Examples of stimuli in which the head was rotated about the pitch axis (left column 45° downwards, middle column 0° and right column 45° upwards) and lighting rotated about the pitch axis (top row 45° top lighting, middle row 0° front lighting and bottom row 45° bottom lighting).

Figure 3.

Mean proportion correct for matching upright faces following yaw head rotations of viewing and lighting angles. Error bars represent ±1 SEM.

Figure 4.

The interaction between lighting and orientation for pitch. Mean proportion correct for matching upright and inverted faces following pitch rotations of lighting angles. Note that lighting is relative to the head so that bottom lighting becomes lighting from above after inversion. Error bars represent ±1 SEM.

Figure 5.

Mean proportion correct for matching upright faces following pitch head rotations of viewing and lighting angles. Error bars represent ±1 SEM.

Figure 6.

Examples of stimuli in which the camera was rotated about the yaw axis (left column 45° leftwards, middle column 0° and right column 45° right wards) and lighting rotated about the yaw axis (top row 45° leftwards, middle row 0° front lighting and bottom row 45° rightwards). Note that because we have now manipulated camera rotation, the layout of images does not match with those in Figure 1.

Figure 7.

Examples of stimuli in which the camera was rotated about the pitch axis (left column 45° below centre, middle column 0° and right column 45° above centre) and lighting rotated about the pitch axis (top row 45° top lighting, middle row 0° front lighting and bottom row 45° bottom lighting). Note that because we have now manipulated camera rotation, the layout of images does not match with those in Figure 2.

Figure 8.

Mean proportion correct for matching upright faces following yaw rotations of camera and lighting angles. Error bars represent ±1 SEM.

Figure 9.

Mean proportion correct for matching upright faces following pitch rotations of camera and lighting angles. Error bars represent ±1 SEM.