EHCTNet: Enhanced Hybrid of CNN and Transformer Network for Remote Sensing Image Change Detection

TL;DR

An enhanced hybrid of CNN and Transformer network (EHCTNet) is proposed for effectively mining the change information of interest and shows that EHCTNet detects more intact and continuous changed areas and perceives more accurate neighboring distinction than state of the art models.

Abstract

Remote sensing (RS) change detection incurs a high cost because of false negatives, which are more costly than false positives. Existing frameworks, struggling to improve the Precision metric to reduce the cost of false positive, still have limitations in focusing on the change of interest, which leads to missed detections and discontinuity issues. This work tackles these issues by enhancing feature learning capabilities and integrating the frequency components of feature information, with a strategy to incrementally boost the Recall value. We propose an enhanced hybrid of CNN and Transformer network (EHCTNet) for effectively mining the change information of interest. Firstly, a dual branch feature extraction module is used to extract the multi scale features of RS images. Secondly, the frequency component of these features is exploited by a refined module I. Thirdly, an enhanced token mining module based on the Kolmogorov Arnold Network is utilized to derive semantic information. Finally, the semantic change information's frequency component, beneficial for final detection, is mined from the refined module II. Extensive experiments validate the effectiveness of EHCTNet in comprehending complex changes of interest. The visualization outcomes show that EHCTNet detects more intact and continuous changed areas and perceives more accurate neighboring distinction than state of the art models.

Figures (8)

• Figure 1: Overall network structure of the EHCTNet.
• Figure 2: Illustration of the HCT branch.
• Figure 3: Illustration of CKSA.
• Figure 4: The heatmaps and change results of our EHCTNet are compared with those of two state-of-the-art (SOTA) approaches (BIT ChenHao2022$\&$ VcT JiangBo2023VcT). The baseline means the feature extraction module of our EHCTNet. The yellow rectangle indicates the presence of missed detections, false detections, and incorrect merging of adjacent change areas.
• Figure 5: The improvement of our HCT on LEVIR-CD dataset over ResNet18.