Perception of material appearance: A comparison between painted and rendered images

Abstract

Painters are masters in replicating the visual appearance of materials. While the perception of material appearance is not yet fully understood, painters seem to have acquired an implicit understanding of the key visual cues that we need to accurately perceive material properties. In this study, we directly compare the perception of material properties in paintings and in renderings by collecting professional realistic paintings of rendered materials. From both type of images, we collect human judgments of material properties and compute a variety of image features that are known to reflect material properties. Our study reveals that, despite important visual differences between the two types of depiction, material properties in paintings and renderings are perceived very similarly and are linked to the same image features. This suggests that we use similar visual cues independently of the medium and that the presence of such cues is sufficient to provide a good appearance perception of the materials.

The Legacy of Willem Beurs ‒ Bridging the Gap between Art and Material Perception

“…After testing the perception of a set of properties for different materials depicted in paintings, they concluded that the mechanism of material perception is independent of the medium of representation, as their results were in agreement with previous studies conducted on photographs (Fleming et al, 2013) and computer renderings (Zhang et al, 2019). A similar conclusion has been reached more recently by Delanoy et al (2021) who found no significant perceptual differences between computer renderings and paintings for the judgment of low-and mid-level material properties. Koenderink and van Doorn (2001) investigated the issue of shading in the case of translucent materials.…”

Effect of material properties on emotion: a virtual reality study

“…To validate this hypothesis, we evaluate the strength of specular reflections using computer vision statistics computed on rendered images of the mugs. More precisely we follow the approach from Di Cicco et al (2019) and Delanoy et al (2021) , illustrated in Figure 10 : for each image we extract the luminance profile which runs through the most luminous and sharp highlights (the position of this line of interest is the same for all images), then we extract the two features evaluated as the most relevant by both authors to characterize the strength of reflections: Michelson contrast ( Michelson, 1927 ) obtained as (Imax-Imin)/(Imax + Imin) and highlights sharpness which corresponds to the mean width of the highlights transitions (see Figure 10 ). Those two features are computed on the 16 mug images and normalized such that they both have a mean of 0 and a standard deviation of 1.…”

“…We then computed the “strength of reflections” predictor proposed by Delanoy et al (2021) which is linear combination of Michelson contrast and highlights sharpness with weights resp. equal to 0.6 and 0.4.…”

In‐the‐wild Material Appearance Editing using Perceptual Attributes

“…Our visual perception of an object is guided by its material properties; but, also involves factors such as geometry [VLD07, HFM16], light conditions [HLM06, KFB10, CK15] or motion [DFY * 11, MLMG19]. To reduce the dimensionality of the problem, previous work have focused on understanding single, high‐level appearance properties like glossiness [PFG00, WAKB09, CK15], translucency [GXZ * 13, XZG * 20, GWA * 15] and softness [SFV20, CDD21]; or draw inspiration from artists' implicit understanding of the key visual cues that guide visual perception [DCWP19, DSMG21]. Recent works suggest that material perception may be driven by complex non‐linear statistics better approximated by highly non‐linear models such as neural networks [FS19, SAF21, DLG * 20, LMS * 19].…”