Feature learning network with transformer for multi-label image classification

Wei Zhou, Peng Dou, Tao Su, Haifeng Hu, Zhijie Zheng

Abstract: Highlights•A novel framework termed FL-Tran is proposed to solve the multi-label image classification task.•A multi-scale fusion mechanism is designed to align high-level features and low-level features to learn multi-scale features.•A spatial attention mechanism based on transformer encoder is developed to capture the salient object features in images.•A feature enhancement and suppression mechanism is proposed to excavate various potential useful features through stage-by-stage suppressing the most salient feature in the feature maps.•Experiments on three publicly available datasets validate the superior performance of the proposed FL-Tran model compared with the state-of-the-art methods.