Intrinsic Image Decomposition using Paradigms
This work addresses the fundamental problem of how a visual agent could learn intrinsic image decomposition without explicit geometric, surface, and illumination models, which is significant for unsupervised learning in computer vision.
This paper tackles the problem of intrinsic image decomposition, mapping images to albedo, without relying on ground truth, rendered data, or human judgments. The method, which uses "paradigms" (fake albedos and shading fields) and a novel smoothing procedure, achieves WHDR scores competitive with supervised methods.
Intrinsic image decomposition is the classical task of mapping image to albedo. The WHDR dataset allows methods to be evaluated by comparing predictions to human judgements ("lighter", "same as", "darker"). The best modern intrinsic image methods learn a map from image to albedo using rendered models and human judgements. This is convenient for practical methods, but cannot explain how a visual agent without geometric, surface and illumination models and a renderer could learn to recover intrinsic images. This paper describes a method that learns intrinsic image decomposition without seeing WHDR annotations, rendered data, or ground truth data. The method relies on paradigms - fake albedos and fake shading fields - together with a novel smoothing procedure that ensures good behavior at short scales on real images. Long scale error is controlled by averaging. Our method achieves WHDR scores competitive with those of strong recent methods allowed to see training WHDR annotations, rendered data, and ground truth data. Because our method is unsupervised, we can compute estimates of the test/train variance of WHDR scores; these are quite large, and it is unsafe to rely small differences in reported WHDR.