1 code implementation • 26 Feb 2022 • Nikos Efthymiadis, Giorgos Tolias, Ondrej Chum
To bridge the domain gap we present a novel augmentation technique that is tailored to the task of learning sketch recognition from a training set of natural images.