ChangeChat: An Interactive Model for Remote Sensing Change Analysis via Multimodal Instruction Tuning

Abstract

Remote sensing (RS) change analysis is vital for monitoring Earth’s dynamic processes by detecting alterations in images over time. Traditional change detection methods excels at identifying pixel-level changes but lacks the ability to contextualize these changes. While recent advancements in change captioning offer natural language descriptions of changes, they do not support interactive, user-specific queries. To address these limitations, we introduce ChangeChat, the first bitemporal vision-language model (VLM) specifically designed for interactive RS change analysis. ChangeChat leverages multimodal instruction tuning to handle complex queries such as change captioning, category-specific quantification, and change localization. To further enhance the model’s capabilities, we developed the ChangeChat-87k dataset, created using a combination of rule-based methods and GPT-assisted techniques. Experimental results demonstrate that ChangeChat provides a comprehensive, interactive solution for RS change analysis. It achieves performance comparable to or surpassing state-of-the-art (SOTA) methods on specific tasks, while significantly outperforming the latest general-domain model, GPT-4. Code and pre-trained weights are available at https://github.com/hanlinwu/ChangeChat.