How to Reduce Change Detection to Semantic Segmentation

TL;DR

This paper introduces a novel approach to change detection by reducing it to semantic segmentation, leveraging existing segmentation networks and a new module for better change type discrimination, achieving state-of-the-art results.

Contribution

The paper proposes a new paradigm that simplifies change detection by converting it into semantic segmentation, and introduces the MTF module for improved change type learning.

Findings

Achieves state-of-the-art performance on change detection benchmarks.

The MTF module effectively extracts and fuses temporal change information.

The proposed C-3PO network serves as a simple, effective baseline for future research.

Abstract

Change detection (CD) aims to identify changes that occur in an image pair taken different times. Prior methods devise specific networks from scratch to predict change masks in pixel-level, and struggle with general segmentation problems. In this paper, we propose a new paradigm that reduces CD to semantic segmentation which means tailoring an existing and powerful semantic segmentation network to solve CD. This new paradigm conveniently enjoys the mainstream semantic segmentation techniques to deal with general segmentation problems in CD. Hence we can concentrate on studying how to detect changes. We propose a novel and importance insight that different change types exist in CD and they should be learned separately. Based on it, we devise a module named MTF to extract the change information and fuse temporal features. MTF enjoys high interpretability and reveals the essential characteristic of CD. And most segmentation networks can be adapted to solve the CD problems with our MTF module. Finally, we propose C-3PO, a network to detect changes at pixel-level. C-3PO achieves state-of-the-art performance without bells and whistles. It is simple but effective and can be considered as a new baseline in this field. Our code is at https://github.com/DoctorKey/C-3PO.