Volume preserving image segmentation with entropy regularized optimal transport and its applications in deep learning

Haifeng Li, Jun Liu*, Li Cui, Haiyang Huang, Xue-Cheng TAI

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

Image segmentation with a volume constraint is an important prior for many real applications. In this work, we present a novel volume preserving image segmentation algorithm, which is based on the entropy and Total Variation (TV) regularized optimal transport theory. The volume and classification constraints can be regarded as two measures preserving constraints in the optimal transport. By studying the dual problem, we develop a simple but efficient dual algorithm for our model. Moreover, to be different from many variational based image segmentation algorithms, the proposed algorithm can be directly unrolled to a new Volume Preserving and TV regularized softmax (VPTV-softmax) layer for semantic segmentation in the popular Deep Convolution Neural Network (DCNN). The experiment results show that our proposed model is very competitive and can improve the performance of many semantic segmentation networks such as the popular U-net and DeepLabv3+.

Original languageEnglish
Article number102845
JournalJournal of Visual Communication and Image Representation
Volume71
DOIs
Publication statusPublished - Aug 2020

Scopus Subject Areas

  • Signal Processing
  • Media Technology
  • Computer Vision and Pattern Recognition
  • Electrical and Electronic Engineering

User-Defined Keywords

  • DCNN
  • Entropic regularization
  • Image segmentation
  • Optimal transport
  • TV regularization
  • Volume preserving

Fingerprint

Dive into the research topics of 'Volume preserving image segmentation with entropy regularized optimal transport and its applications in deep learning'. Together they form a unique fingerprint.

Cite this