Object-Guided Instance Segmentation for Biological Images

Authors

  • Jingru Yi Rutgers University
  • Hui Tang Tecent Medical AI Lab
  • Pengxiang Wu Rutgers University
  • Bo Liu Rutgers University
  • Daniel J. Hoeppner Lieber Institute for Brain Development
  • Dimitris N. Metaxas Rutgers University
  • Lianyi Han Tecent Medical AI Lab
  • Wei Fan Tecent Medical AI Lab

DOI:

https://doi.org/10.1609/aaai.v34i07.6960

Abstract

Instance segmentation of biological images is essential for studying object behaviors and properties. The challenges, such as clustering, occlusion, and adhesion problems of the objects, make instance segmentation a non-trivial task. Current box-free instance segmentation methods typically rely on local pixel-level information. Due to a lack of global object view, these methods are prone to over- or under-segmentation. On the contrary, the box-based instance segmentation methods incorporate object detection into the segmentation, performing better in identifying the individual instances. In this paper, we propose a new box-based instance segmentation method. Mainly, we locate the object bounding boxes from their center points. The object features are subsequently reused in the segmentation branch as a guide to separate the clustered instances within an RoI patch. Along with the instance normalization, the model is able to recover the target object distribution and suppress the distribution of neighboring attached objects. Consequently, the proposed model performs excellently in segmenting the clustered objects while retaining the target object details. The proposed method achieves state-of-the-art performances on three biological datasets: cell nuclei, plant phenotyping dataset, and neural cells.

Downloads

Published

2020-04-03

How to Cite

Yi, J., Tang, H., Wu, P., Liu, B., Hoeppner, D. J., Metaxas, D. N., Han, L., & Fan, W. (2020). Object-Guided Instance Segmentation for Biological Images. Proceedings of the AAAI Conference on Artificial Intelligence, 34(07), 12677-12684. https://doi.org/10.1609/aaai.v34i07.6960

Issue

Section

AAAI Technical Track: Vision