{"corpus_id":255633734,"paper_sha":"bbf3a2aba68c41c12eeb24cf85afbc38eb111b16","doi":"10.1109/LGRS.2023.3233979","arxiv_id":null,"pmid":null,"pmcid":null,"mag_id":null,"dblp_id":"journals/lgrs/HeDFMZ23","acl_id":null,"title":"Dual-Range Context Aggregation for Efficient Semantic Segmentation in Remote Sensing Images","year":2023,"publication_date":null,"venue":"IEEE Geoscience and Remote Sensing Letters","journal":{"name":"IEEE Geoscience and Remote Sensing Letters","pages":"1-5","volume":"20"},"journal_issn":null,"journal_title":null,"publication_types":["JournalArticle"],"pubmed_pub_types":null,"s2_fields_of_study":["Computer Science","Environmental Science"],"reference_count":34,"citation_count":8,"influential_citation_count":0,"is_open_access":false,"arxiv_categories":null,"arxiv_license":null,"arxiv_journal_ref":null,"mesh_headings":null,"chemicals":null,"comments_corrections":null,"source_flags":1,"s2_open_access_pdf_url":null,"s2_open_access_landing_url":null,"s2_open_access_license":null,"s2_open_access_status":null,"pmc_open_access_pdf_url":null,"pmc_open_access_landing_url":null,"pmc_open_access_license":null,"pmc_open_access_status":null,"unpaywall_open_access_pdf_url":null,"unpaywall_open_access_landing_url":null,"unpaywall_open_access_license":null,"unpaywall_open_access_status":null,"abstract":"Although introducing self-attention mechanisms is beneficial to establish long-range dependencies and explore global context information in the task of remote sensing image semantic segmentation, it results in expensive computation and large memory cost. In this letter, we address this dilemma by proposing a lightweight dual-range context aggregation network (LDCANet) for efficient remote sensing image semantic segmentation. First, a dual-range context aggregation module (DCAM) is designed to aggregate the local features and the global semantic context acquired by convolutions and self-attention, respectively, where self-attention is implemented easily by applying two cascaded linear layers to reduce the computational complexity. Furthermore, a simple and lightweight decoder is employed to combine information from different levels, in which a multilayer perceptron (MLP)-based efficient linear block (ELB) is proposed to yield a strong and efficient representation. Experiments conducted on the International Society for Photogrammetry and Remote Sensing (ISPRS) Vaihingen dataset and the Gaofen Image dataset (GID) prove that our LDCANet achieves an excellent trade-off between segmentation accuracy and computational efficiency. In particular, our method achieves 74.12% mean intersection over union (mIoU) on the ISPRS Vaihingen dataset and 61.42% mIoU on the GID with only 4.98-M parameter size.","claims":[{"public_id":"cl_42a224eb5f907ca78869f279becb010f","status":"active","text":"A dual-range context aggregation module (DCAM) aggregates local features and global semantic context acquired by convolutions and self-attention, where self-attention is implemented by two cascaded linear layers to reduce computational complexity.","confidence":0.9,"contributors":[{"id":32,"public_id":"7c402c1b98","public_label":"뀨 (7c402c1b98)","roles":["extraction"],"url":"https://sah.borca.ai/u/7c402c1b98"},{"id":127,"public_id":"yy3835ars5","public_label":"Anonymous (yy3835ars5)","roles":["review"],"url":"https://sah.borca.ai/u/yy3835ars5"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/claims/cl_42a224eb5f907ca78869f279becb010f"},{"public_id":"cl_04dc4873d8237b7df51fc133de81bcda","status":"active","text":"A multilayer perceptron (MLP)-based efficient linear block (ELB) is proposed in a lightweight decoder to combine information from different levels for strong and efficient representation.","confidence":0.9,"contributors":[{"id":32,"public_id":"7c402c1b98","public_label":"뀨 (7c402c1b98)","roles":["extraction"],"url":"https://sah.borca.ai/u/7c402c1b98"},{"id":127,"public_id":"yy3835ars5","public_label":"Anonymous (yy3835ars5)","roles":["review"],"url":"https://sah.borca.ai/u/yy3835ars5"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/claims/cl_04dc4873d8237b7df51fc133de81bcda"},{"public_id":"cl_760b64c3e8a8713780e8b2db83f79415","status":"active","text":"LDCANet achieves 74.12% mIoU on the ISPRS Vaihingen dataset and 61.42% mIoU on the GID with only 4.98-M parameter size.","confidence":0.95,"contributors":[{"id":32,"public_id":"7c402c1b98","public_label":"뀨 (7c402c1b98)","roles":["extraction"],"url":"https://sah.borca.ai/u/7c402c1b98"},{"id":127,"public_id":"yy3835ars5","public_label":"Anonymous (yy3835ars5)","roles":["review"],"url":"https://sah.borca.ai/u/yy3835ars5"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/claims/cl_760b64c3e8a8713780e8b2db83f79415"},{"public_id":"cl_86aa5a49b14e899067687caef504a4ea","status":"active","text":"LDCANet achieves an excellent trade-off between segmentation accuracy and computational efficiency on remote sensing image semantic segmentation.","confidence":0.85,"contributors":[{"id":32,"public_id":"7c402c1b98","public_label":"뀨 (7c402c1b98)","roles":["extraction"],"url":"https://sah.borca.ai/u/7c402c1b98"},{"id":127,"public_id":"yy3835ars5","public_label":"Anonymous (yy3835ars5)","roles":["review"],"url":"https://sah.borca.ai/u/yy3835ars5"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/claims/cl_86aa5a49b14e899067687caef504a4ea"}],"concepts":[{"public_id":"co_10942c820a10fcd597a048fa2b7ece38","status":"active","name":"self-attention","description":"Mechanism used in DCAM to capture global context, implemented with two cascaded linear layers for efficiency.","types":["mechanism"],"aliases":[],"contributors":[{"id":32,"public_id":"7c402c1b98","public_label":"뀨 (7c402c1b98)","roles":["extraction"],"url":"https://sah.borca.ai/u/7c402c1b98"},{"id":127,"public_id":"yy3835ars5","public_label":"Anonymous (yy3835ars5)","roles":["review"],"url":"https://sah.borca.ai/u/yy3835ars5"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/concepts/co_10942c820a10fcd597a048fa2b7ece38"},{"public_id":"co_10f529442523f4da10a869477635b893","status":"active","name":"Gaofen Image dataset (GID)","description":"Remote sensing image dataset from Gaofen satellites used for evaluation.","types":["dataset"],"aliases":["GID"],"contributors":[{"id":32,"public_id":"7c402c1b98","public_label":"뀨 (7c402c1b98)","roles":["extraction"],"url":"https://sah.borca.ai/u/7c402c1b98"},{"id":127,"public_id":"yy3835ars5","public_label":"Anonymous (yy3835ars5)","roles":["review"],"url":"https://sah.borca.ai/u/yy3835ars5"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/concepts/co_10f529442523f4da10a869477635b893"},{"public_id":"co_3dfb4d2a751f7ebedcc4464aa75a6cfe","status":"active","name":"mean intersection over union (mIoU)","description":"Metric used to measure segmentation accuracy in the experiments.","types":["metric"],"aliases":["mIoU"],"contributors":[{"id":32,"public_id":"7c402c1b98","public_label":"뀨 (7c402c1b98)","roles":["extraction"],"url":"https://sah.borca.ai/u/7c402c1b98"},{"id":127,"public_id":"yy3835ars5","public_label":"Anonymous (yy3835ars5)","roles":["review"],"url":"https://sah.borca.ai/u/yy3835ars5"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/concepts/co_3dfb4d2a751f7ebedcc4464aa75a6cfe"},{"public_id":"co_a1cc123960528e69072e0f1de4b2aed9","status":"active","name":"cascaded linear layers","description":"Two linear layers applied in sequence to implement self-attention with reduced computational complexity.","types":["component"],"aliases":[],"contributors":[{"id":32,"public_id":"7c402c1b98","public_label":"뀨 (7c402c1b98)","roles":["extraction"],"url":"https://sah.borca.ai/u/7c402c1b98"},{"id":127,"public_id":"yy3835ars5","public_label":"Anonymous (yy3835ars5)","roles":["review"],"url":"https://sah.borca.ai/u/yy3835ars5"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/concepts/co_a1cc123960528e69072e0f1de4b2aed9"},{"public_id":"co_b1b867dfa1c3c227ff9c83585bd943bb","status":"active","name":"lightweight decoder","description":"Simple and lightweight decoder that combines information from different levels for segmentation.","types":["component"],"aliases":[],"contributors":[{"id":32,"public_id":"7c402c1b98","public_label":"뀨 (7c402c1b98)","roles":["extraction"],"url":"https://sah.borca.ai/u/7c402c1b98"},{"id":127,"public_id":"yy3835ars5","public_label":"Anonymous (yy3835ars5)","roles":["review"],"url":"https://sah.borca.ai/u/yy3835ars5"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/concepts/co_b1b867dfa1c3c227ff9c83585bd943bb"},{"public_id":"co_b8c7d07d8dc54b5f78d12d92f9a0a62f","status":"active","name":"multilayer perceptron (MLP)-based efficient linear block (ELB)","description":"Proposed block in the lightweight decoder that yields strong and efficient representation using MLP.","types":["component"],"aliases":["ELB"],"contributors":[{"id":32,"public_id":"7c402c1b98","public_label":"뀨 (7c402c1b98)","roles":["extraction"],"url":"https://sah.borca.ai/u/7c402c1b98"},{"id":127,"public_id":"yy3835ars5","public_label":"Anonymous (yy3835ars5)","roles":["review"],"url":"https://sah.borca.ai/u/yy3835ars5"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/concepts/co_b8c7d07d8dc54b5f78d12d92f9a0a62f"},{"public_id":"co_bd86b289608b282b65b40e1d562a7188","status":"active","name":"dual-range context aggregation module (DCAM)","description":"Module designed to aggregate local features and global semantic context using convolutions and self-attention.","types":["component"],"aliases":["DCAM"],"contributors":[{"id":32,"public_id":"7c402c1b98","public_label":"뀨 (7c402c1b98)","roles":["extraction"],"url":"https://sah.borca.ai/u/7c402c1b98"},{"id":127,"public_id":"yy3835ars5","public_label":"Anonymous (yy3835ars5)","roles":["review"],"url":"https://sah.borca.ai/u/yy3835ars5"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/concepts/co_bd86b289608b282b65b40e1d562a7188"},{"public_id":"co_c2339094632f038e7294c47dc849f609","status":"active","name":"remote sensing image semantic segmentation","description":"Task of assigning semantic labels to pixels in remote sensing images, targeted by LDCANet.","types":["task"],"aliases":[],"contributors":[{"id":32,"public_id":"7c402c1b98","public_label":"뀨 (7c402c1b98)","roles":["extraction"],"url":"https://sah.borca.ai/u/7c402c1b98"},{"id":127,"public_id":"yy3835ars5","public_label":"Anonymous (yy3835ars5)","roles":["review"],"url":"https://sah.borca.ai/u/yy3835ars5"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/concepts/co_c2339094632f038e7294c47dc849f609"},{"public_id":"co_c75c6875dff238021b33773d642a5123","status":"active","name":"ISPRS Vaihingen dataset","description":"Remote sensing image dataset from the International Society for Photogrammetry and Remote Sensing used for evaluation.","types":["dataset"],"aliases":[],"contributors":[{"id":32,"public_id":"7c402c1b98","public_label":"뀨 (7c402c1b98)","roles":["extraction"],"url":"https://sah.borca.ai/u/7c402c1b98"},{"id":127,"public_id":"yy3835ars5","public_label":"Anonymous (yy3835ars5)","roles":["review"],"url":"https://sah.borca.ai/u/yy3835ars5"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/concepts/co_c75c6875dff238021b33773d642a5123"},{"public_id":"co_f3539afdc0ffda7143352b5d9890a2c8","status":"active","name":"LDCANet","description":"Lightweight dual-range context aggregation network proposed for efficient remote sensing image semantic segmentation.","types":["method","model"],"aliases":[],"contributors":[{"id":32,"public_id":"7c402c1b98","public_label":"뀨 (7c402c1b98)","roles":["extraction"],"url":"https://sah.borca.ai/u/7c402c1b98"},{"id":127,"public_id":"yy3835ars5","public_label":"Anonymous (yy3835ars5)","roles":["review"],"url":"https://sah.borca.ai/u/yy3835ars5"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/concepts/co_f3539afdc0ffda7143352b5d9890a2c8"}],"external_ids":{"DOI":"10.1109/LGRS.2023.3233979","ArXiv":null,"PubMed":null,"PubMedCentral":null,"MAG":null,"DBLP":"journals/lgrs/HeDFMZ23","ACL":null},"open_access":{"is_open_access":false,"pdf_url":null,"landing_url":"https://sah.borca.ai/papers/255633734","source":null,"pdf_url_source":null,"license":null,"reason":"pdf_url_not_indexed"},"reference_availability":{"status":"available","references_indexed":true,"full_text_available":false,"full_text_source":null,"count_basis":"semantic_scholar_metadata","extraction_status":"not_applicable","reason":null},"source":{"provider":"episteme2","base_corpus":"semantic_scholar_dump","freshness_mode":"unknown","basis":["semantic_scholar_metadata","postgres_metadata"],"limits":["paper metadata is based on indexed upstream scholarly datasets","claims and concepts are available only for extracted papers","absence of claims or concepts means no extracted graph data is available in this response"],"status":"available","degraded":false,"degraded_reasons":[],"diagnostics":{"status":"available","degraded":false,"degraded_reasons":[],"metadata_status":"available","graph_status":"available","abstract_status":"available"},"source_flags":1},"paper_id":638180,"paper_uid":"92a59729-58fe-4c9a-a236-5e36bd7fc222","canonical_identity":{"paper_id":638180,"paper_uid":"92a59729-58fe-4c9a-a236-5e36bd7fc222","identity_status":"available","lookup_basis":"semantic_scholar_external_id","compatibility_path":"corpus_id"},"url":"https://sah.borca.ai/papers/255633734"}