Referring Remote Sensing Image Segmentation with Cross-view Semantics Interaction Network

TL;DR

This paper introduces CSINet, a novel unified framework for remote sensing image segmentation that leverages cross-view semantics interaction to improve accuracy on targets of varying scales, especially tiny or ambiguous ones.

Contribution

The paper proposes a new parallel, unified segmentation network with cross-view semantics interaction, including CVWin modules and a CDAD decoder, to better handle scale variation in remote sensing images.

Findings

Significant performance improvements over existing methods.

Effective handling of tiny and ambiguous targets.

Maintains high speed while enhancing global and local semantics.

Abstract

Recently, Referring Remote Sensing Image Segmentation (RRSIS) has aroused wide attention. To handle drastic scale variation of remote targets, existing methods only use the full image as input and nest the saliency-preferring techniques of cross-scale information interaction into traditional single-view structure. Although effective for visually salient targets, they still struggle in handling tiny, ambiguous ones in lots of real scenarios. In this work, we instead propose a paralleled yet unified segmentation framework Cross-view Semantics Interaction Network (CSINet) to solve the limitations. Motivated by human behavior in observing targets of interest, the network orchestrates visual cues from remote and close distances to conduct synergistic prediction. In its every encoding stage, a Cross-View Window-attention module (CVWin) is utilized to supplement global and local semantics into close-view and remote-view branch features, finally promoting the unified representation of feature in every encoding stage. In addition, we develop a Collaboratively Dilated Attention enhanced Decoder (CDAD) to mine the orientation property of target and meanwhile integrate cross-view multiscale features. The proposed network seamlessly enhances the exploitation of global and local semantics, achieving significant improvements over others while maintaining satisfactory speed.