Abstract
An often overlooked but integral aspect of object recognition is material perception, e.g., whether an object is made of plastic, pearl, satin, or steel. There is a growing body of work investigating the visual perception of the properties of materials, with a particular focus on gloss. For example, the appearance of specular reflections (e.g., their size, sharpness, and contrast) has been shown to affect how glossy a surface looks. Yet little is known about how we recognise different material classes, nor the precise relationship between material properties like gloss and material class. To investigate this, we rendered complex glossy objects in natural illumination fields. Stimuli varied in shape, illumination conditions, diffuse shading, and specular reflectance parameters. Separate sets of participants judged the glossiness and material category of each object. We found that manipulating an object’s surface reflectance properties not only changed perceived gloss (i.e., how glossy is it?), but it also led to qualitative changes in perceived material class (i.e., what material is it?). We measured and then causally manipulated visual image features integral to gloss perception and found that these features also predicted changes in material appearance. Surprisingly, while linear combinations of the image features predicted perceived gloss, the strength of this linear relationship differed within each material category. This suggests that material class mediates the processing of image features for perceived gloss. Our results therefore do not support a traditional feedforward view that assumes that material perception proceeds from low-level image measurements to mid-level estimates of surface properties to high-level material classes. Instead, we suggest that the perception and neural processing of material properties like surface gloss should be considered in the context of material recognition, and more generally that the image structure that triggers our perception of surface gloss also plays an important role in visual categorisation.