Region Proposal Rectification Towards Robust Instance Segmentation of Biological Images

• Published in: Medical Image Computing and Computer Assisted Intervention – MICCAI 2022 pp. 129 - 139
• Main Authors: Zhangli, Qilong, Yi, Jingru, Liu, Di, He, Xiaoxiao, Xia, Zhaoyang, Chang, Qi, Han, Ligong, Gao, Yunhe, Wen, Song, Tang, Haiming, Wang, He, Zhou, Mu, Metaxas, Dimitris
• Format: Book Chapter
• Language: English
• Published: Cham Springer Nature Switzerland
• Series: Lecture Notes in Computer Science
• Subjects: Cell; Detection; Instance segmentation; Pathology
• ISBN: 9783031164392; 3031164393
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-031-16440-8_13

Top-down instance segmentation framework has shown its superiority in object detection compared to the bottom-up framework. While it is efficient in addressing over-segmentation, top-down instance segmentation suffers from over-crop problem. However, a complete segmentation mask is crucial for biological image analysis as it delivers important morphological properties such as shapes and volumes. In this paper, we propose a region proposal rectification (RPR) module to address this challenging incomplete segmentation problem. In particular, we offer a progressive ROIAlign module to introduce neighbor information into a series of ROIs gradually. The ROI features are fed into an attentive feed-forward network (FFN) for proposal box regression. With additional neighbor information, the proposed RPR module shows significant improvement in correction of region proposal locations and thereby exhibits favorable instance segmentation performances on three biological image datasets compared to state-of-the-art baseline methods. Experimental results demonstrate that the proposed RPR module is effective in both anchor-based and anchor-free top-down instance segmentation approaches, suggesting the proposed method can be applied to general top-down instance segmentation of biological images. Code is available (https://github.com/qzhangli/RPR).