Abstract

Many Computer Vision problems arise from information processing of data sources with nuisance variances like scale, orientation, contrast, perspective foreshortening or – in medical imaging – staining and local warping. In most cases these variances can be stated a priori and can be used to improve the generalization of recognition algorithms. We propose a novel supervised feature learning approach, which effificiently extracts information from these constraints to produce interpretable, transformation-invariant features. The proposed method can incorporate a large class of transformations, e.g., shifts, rotations, change of scale, morphological operations, non-linear distortions, photometric transformations, etc. These features boost the discrimination power of a novel image classifification and segmentation method, which we call Transformation-Invariant Convolutional Jungles (TICJ). We test the algorithm on two benchmarks in face recognition and medical imaging, where it achieves state of the art results, while being computationally signifificantly more effificient than Deep Neural Networks.