Multimodal　change　detection　(CD)　is　a　practical　but　challenging　task　in　remote　sensing.　Multimodal　change　detection　allows　remote　sensing　images　to　be　acquired　by　different　sensors,　which　is　important　for　disaster　damage　assessment.　In　this　study,　a　multimodal　information　exchange　network　(MIENet)　is　explored　by　abandoning　existing　feature　space　transformation　methods　for　solving　multimodal　image　heterogeneity.　The　network　incorporates　an　information　exchange　module　(IEM)　in　the　feature　extraction　layer,　utilizing　an　attention　mechanism　to　facilitate　the　exchange　of　information　between　different　modalities.　Through　this　module,　we　can　mutually　learn　and　perceive　the　contextual　information　of　different　modality　features,　achieving　domain　adaptation　between　different　modality　domains　to　a　certain　extent.　Extensive　experiments　on　the　California　dataset　demonstrate　that　MIENet　outperforms　existing　feature　space　transformation　methods　in　recognizing　change　areas　in　multimodal　images.　Compared　to　other　methods　applied　to　this　dataset,　MIENet　achieved　the　highest　F1　scores　and　MIOU　values:　90.35%　and　89.69%.　These　results　validate　the　effectiveness　of　our　proposed　MIENet　method　in　multimodal　change　detection.　©　2024　Copyright　held　by　the　owner/author(s).