OCMR: A comprehensive framework for optical chemical molecular recognition

Yan Wang, Ruochi Zhang, Shengde Zhang, Liming Guo, Qiong Zhou, Bowen Zhao, Xiaotong Mo, Qian Yang, Yajuan Huang, Kewei Li, Yusi Fan, Lan Huang, Fengfeng Zhou

Abstract: Highlights•Optical chemical structure recognition (OCSR) from an image is a challenging task.•The Only Chemical Structure Recognition (OCMR) framework is proposed for OCSR.•OCMR uses multiple refined neural networks for the individual OCSR sub-tasks.•OCMR outperforms the state-of-the-art results on multiple benchmark datasets.•The code and in-house collected benchmark dataset are released.