Transformer-based Multimodal Change Detection with Multitask Consistency Constraints

TL;DR

This paper introduces a Transformer-based multimodal change detection method that effectively combines digital surface models and aerial images, addressing multitask conflicts with a consistency constraint to improve detection accuracy.

Contribution

It proposes a novel Transformer network with a consistency constraint for multimodal change detection, leveraging cross-attention and pseudo-changes to unify semantic and height change tasks.

Findings

Outperforms five state-of-the-art methods in multitask change detection

Constructed a new DSM-to-image dataset from three Dutch cities

Demonstrates the effectiveness of the consistency constraint in improving results

Abstract

Change detection plays a fundamental role in Earth observation for analyzing temporal iterations over time. However, recent studies have largely neglected the utilization of multimodal data that presents significant practical and technical advantages compared to single-modal approaches. This research focuses on leveraging {pre-event} digital surface model (DSM) data and {post-event} digital aerial images captured at different times for detecting change beyond 2D. We observe that the current change detection methods struggle with the multitask conflicts between semantic and height change detection tasks. To address this challenge, we propose an efficient Transformer-based network that learns shared representation between cross-dimensional inputs through cross-attention. {It adopts a consistency constraint to establish the multimodal relationship. Initially, pseudo-changes are derived by employing height change thresholding. Subsequently, the $L2$ distance between semantic and pseudo-changes within their overlapping regions is minimized. This explicitly endows the height change detection (regression task) and semantic change detection (classification task) with representation consistency.} A DSM-to-image multimodal dataset encompassing three cities in the Netherlands was constructed. It lays a new foundation for beyond-2D change detection from cross-dimensional inputs. Compared to five state-of-the-art change detection methods, our model demonstrates consistent multitask superiority in terms of semantic and height change detection. Furthermore, the consistency strategy can be seamlessly adapted to the other methods, yielding promising improvements.