Introduction

We address the problem of joint detection and segmentation of multiple object instances in an image, which is a key step towards scene understanding. In particular, we aim to reduce the annotation cost and integrate prior knowledge into data-driven methods. We have proposed several active inference and efficient learning strategies, including weakly supervised semantic segmentation, dynamic model estimation, semi-supervised learning and exemplar-based approach to the task of instance segmentation。

Related Works

Shape-aware Semi-supervised 3D Semantic Segmentation for Medical Images Shuailin Li, Chuyu Zhang, Xuming He International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), 2020

LatentGNN: Learning Efficient Non-local Relations for Visual Recognition Songyang Zhang, Shipeng Yan, Xuming He International Conference on Machine Learning (ICML),2019

Dynamic Context Correspondence Network for Semantic Alignment Shuaiyi Huang, Qiuyue Wang, Songyang Zhang, Shipeng Yan, Xuming He International Conference on Computer Vision (ICCV), 2019