nehay06/CrossDomainImageMatching

Images of same objects across different visual domains like sketches, paintings, photographs in varied environment, tend to show large difference at raw pixel level.This makes it difficult to match such images. We propose a simple and efficient method to compute cross domain image similarity by utilizing standard object detection framework to find relative importance of features. These weighted features are then used to find image matches, based on visual and contextual similarity. For robust experimentation analysis, we test our approach on UFU-DDD[6], PASCALVOC[ 18], INRIA Holidays[19], Random Flickr images[20] and Sketchy datasets[21].