“Javitz sharply criticized his ideas, cautioning that his approach required a subjective appraisal riddled with personal aesthetic bias that would endanger the objective, impartial study of images.”
This was funny to me because if one visits the NYPL Picture Collection, the categories are quite subjective. For example, in an image of the moon in the night sky, how does one determine if the picture should fall under moon, or sky?
The two genealogies of image classification discussed in Kamin’s piece also made me think of the different computer vision techniques used in analyzing images and if they might fall into either category. For example, I might categorize object detection under Javitz’s line of thought, whereas something like the watershed algorithm (which views the image as almost a black and white topographical map seeing brighter areas as elevated points) better fitting with Karpel’s philosophy. Regardless of this binary categorization, I think the notion that there are many, many ways to analyze an image is interesting and also carries over into Computer Vision.