GMC: A general framework of multi-stage context learning and utilization for visual detection tasks

Xuan Wang, Hao Tang, Zhigang Zhu

Abstract: Highlights•A general framework for multistage context learning and utilization in computer vision.•Applicable to various visual detection tasks.•An automation mechanism for minimal modification in using different deep learning models.•Capability of different types of context information (local, semantic, spatial) at various stages of the detection process.