AGNet: Automatic generation network for skin imaging reports

Fan Wu, Haiqiong Yang, Linlin Peng, Zongkai Lian, Mingxin Li, Gang Qu, Shancheng Jiang, Yu Han

Abstract: Highlights•This research focuses on the automatic generation of reports based on images of skin pathologies.•There are three principal challenges when generating such reports: accuracy, efficiency, and visualization.•We propose AGNet, a novel and joint deep learning model to solve above challenges.•For the attention module, we propose a “similarity-attention” mechanism.