Abstract

We investigate the problem of fine-grained sketch-basedimage retrieval (SBIR), where free-hand human sketches areused as queries to perform instance-level retrieval of images. This is an extremely challenging task because (i) vi-sual comparisons not only need to be fine-grained but also executed cross-domain, (ii) free-hand (finger) sketches arehighly abstract, making fine-grained matching harder, andmost importantly (iii) annotated cross-domain sketch-photodatasets required for training are scarce, challenging manystate-of-the-art machine learning techniques. In this paper, for the first time, we address all thesechallenges, providing a step towards the capabilities thatwould underpin a commercial sketch-based image retrieval application. We introduce a new database of 1,432 sketchphoto pairs from two categories with 32,000 fine-grained triplet ranking annotations. We then develop a deep tripletranking model for instance-level SBIR with a novel data augmentation and staged pre-training strategy to alleviate the issue of insufficient fine-grained training data. Extensive experiments are carried out to contribute a variety of insights into the challenges of data sufficiency and over-fitting avoidance when training deep networks for finegrained cross-domain ranking tasks.