In the past few years, a handful of scholars worldwide trained in computer vision, pattern recognition, image processing and art history have applied the techniques of computer vision and computer graphics to problems in the history and interpretation of art. These new computer methods, guided by art historical knowledge, are shedding new light on art works, artistic praxis, and more. Again and again, we see that for some problems these computer methods are more sensitive, more "perceptive," than even a trained artist or art historian, at least for a handful of problems. For instance, visual psychologists have shown that most of us—trained art scholars and artists included—are not particularly good at judging perspective or the location of illumination in a photograph, and, by extension, in a painting but these new computer methods can be extremely good at just such tasks. Likewise, computer image processing methods can detect the subtlest variations in brush strokes, variations that elude most trained eyes. The computer methods do not supplant connoisseurship, of course, but enhance it, much like a microscope empowers a biologist. Moreover, some highly sophisticated methods, such as fractal analysis of Jackson Pollock's drip paintings, introduce new visual measures never considered by the art community. Computer methods also dewarp distorted images in curved mirrors depicted within paintings and thereby provide new views into the artists' studios. Finally, computer graphics reconstructions of artists' studios allow scholars to explore "what if" scenarios and thus better understand the working methods of some artists. As such, conservators, curators and art historians may find these computer methods to be valuable tools, once the strengths and limitations of these methods are fully understood.