BINDER: On Identification and Retrieval of Near-Duplicate Biological Images: a New Dataset and Protocol

Abstract: Manipulation and re-use of images in scientific publications is a growing issue, not only for biomedical publishers but also for the research community in general. In this work, we introduce BINDER – Bio-Image Near-Duplicate Examples Repository, a novel dataset to help researchers develop, train, and test models to detect same-source biomedical images. BINDER contains 7,490 unique image patches – the training set – as well as 1,821 patches – split in validation and test sets – with accompanying manipulations obtained by means of transformations including rotation, translation, scale, perspective transform, contrast adjustment, and/or compression artifacts. In addition, we show how novel adaptations of existing image retrieval and metric learning models, when trained on this dataset, can be applied to achieve high-accuracy inference results, creating a baseline for future work. In aggregate, we thus present a supervised protocol for near-duplicate image identification and retrieval without any “real-world” training example. This is part of the Image Forensics project: Image Forensics