Paul, DipanjyotiChowdhury, ArpitaXiong, XinqiChang, Feng-JuCarlyn, DavidStevens, SamuelProvost, KaiyaKarpatne, AnujCarstens, BryanRubenstein, Daniel I.Stewart, Charles V.Berger-Wolf, Tanya Y.Su, YuChao, Wei-Lun2024-02-272024-02-272023https://hdl.handle.net/10919/118170We present a novel usage of Transformers to make image classification interpretable. Unlike mainstream classifiers that wait until the last fully-connected layer to incorporate class information to make predictions, we investigate a proactive approach, asking each class to search for itself in an image. We realize this idea via a Transformer encoder-decoder inspired by DEtection TRansformer (DETR). We learn “class-specific” queries (one for each class) as input to the decoder, enabling each class to localize its patterns in an image via cross-attention. We name our approach INterpretable TRansformer (INTR), which is fairly easy to implement and exhibits several compelling properties. We show that INTR intrinsically encourages each class to attend distinctively; the cross-attention weights thus provide a faithful interpretation of the prediction. Interestingly, via “multi-head” cross-attention, INTR could identify different “attributes” of a class, making it particularly suitable for fine-grained classification and analysis, which we demonstrate on eight datasets. Our code and pre-trained model are publicly accessible at https://github.com/Imageomics/INTR.application/pdfenIn CopyrightArticleabs/2311.04157Karpatne, Anuj [0000-0003-1647-3534]