{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/recalibrating-fully-convolutional-networks","title":"Recalibrating Fully Convolutional Networks with Spatial and Channel 'Squeeze & Excitation' Blocks","arxiv_id":"1808.08127","date":"2018-08-23","proceeding":null,"authors":["Abhijit Guha Roy","Nassir Navab","Christian Wachinger"],"abstract":"In a wide range of semantic segmentation tasks, fully convolutional neural\nnetworks (F-CNNs) have been successfully leveraged to achieve state-of-the-art\nperformance. Architectural innovations of F-CNNs have mainly been on improving\nspatial encoding or network connectivity to aid gradient flow. In this article,\nwe aim towards an alternate direction of recalibrating the learned feature maps\nadaptively; boosting meaningful features while suppressing weak ones. The\nrecalibration is achieved by simple computational blocks that can be easily\nintegrated in F-CNNs architectures. We draw our inspiration from the recently\nproposed 'squeeze & excitation' (SE) modules for channel recalibration for\nimage classification. Towards this end, we introduce three variants of SE\nmodules for segmentation, (i) squeezing spatially and exciting channel-wise,\n(ii) squeezing channel-wise and exciting spatially and (iii) joint spatial and\nchannel 'squeeze & excitation'. We effectively incorporate the proposed SE\nblocks in three state-of-the-art F-CNNs and demonstrate a consistent\nimprovement of segmentation accuracy on three challenging benchmark datasets.\nImportantly, SE blocks only lead to a minimal increase in model complexity of\nabout 1.5%, while the Dice score increases by 4-9% in the case of U-Net. Hence,\nwe believe that SE blocks can be an integral part of future F-CNN\narchitectures.","url_abs":"http://arxiv.org/abs/1808.08127v1","url_pdf":"http://arxiv.org/pdf/1808.08127v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"recalibrating-fully-convolutional-networks","repo_url":"https://github.com/abhi4ssj/squeeze_and_excitation","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"recalibrating-fully-convolutional-networks","repo_url":"https://github.com/ai-med/squeeze_and_excitation","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"recalibrating-fully-convolutional-networks","repo_url":"https://github.com/highwaywu/tianchi-fft2","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"recalibrating-fully-convolutional-networks","repo_url":"https://github.com/ioanvl/1d_squeeze_excitation","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"recalibrating-fully-convolutional-networks","repo_url":"https://github.com/qubvel/segmentation_models.pytorch/blob/a6e1123983548be55d4d1320e0a2f5fd9174d4ac/segmentation_models_pytorch/base/modules.py","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"segmentation","task_name":"Segmentation"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"},{"task_slug":"image-classification","task_name":"image-classification"}],"methods":[{"method_slug":"average-pooling","method_name":"Average Pooling"},{"method_slug":"channel-squeeze-and-spatial-excitation","method_name":"Channel Squeeze and Spatial Excitation"},{"method_slug":"concatenated-skip-connection","method_name":"Concatenated Skip Connection"},{"method_slug":"scse","method_name":"Concurrent Spatial and Channel Squeeze & Excitation"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"max-pooling","method_name":"Max Pooling"},{"method_slug":"relu","method_name":"ReLU"},{"method_slug":"sigmoid-activation","method_name":"Sigmoid Activation"},{"method_slug":"squeeze-and-excitation-block","method_name":"Squeeze-and-Excitation Block"},{"method_slug":"u-net","method_name":"U-Net"},{"method_slug":"scse-1","method_name":"scSE"}],"datasets_introduced":[],"methods_introduced":[{"slug":"channel-squeeze-and-spatial-excitation","name":"Channel Squeeze and Spatial Excitation","full_name":"Channel Squeeze and Spatial Excitation (sSE)"},{"slug":"scse","name":"Concurrent Spatial and Channel Squeeze & Excitation","full_name":"Concurrent Spatial and Channel Squeeze & Excitation (scSE)"},{"slug":"scse-1","name":"scSE","full_name":"Spatial and Channel SE Blocks"}],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1808.08127","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1808.08127"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/ai-med/squeeze_and_excitation","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/ioanvl/1d_squeeze_excitation","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/abhi4ssj/squeeze_and_excitation","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/highwaywu/tianchi-fft2","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/qubvel/segmentation_models.pytorch/blob/a6e1123983548be55d4d1320e0a2f5fd9174d4ac/segmentation_models_pytorch/base/modules.py","reach":null}],"summary":{"ran_draft_wrong":2,"unverified":6},"by_repo_kind":{"listed":{"samples":8,"ran":2,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"f15a49337e69e937","entry":"round_filters","repo":"qubvel/segmentation_models.pytorch","repo_kind":"listed","path":"segmentation_models_pytorch/encoders/_efficientnet.py","file_url":"https://github.com/qubvel/segmentation_models.pytorch/blob/HEAD/segmentation_models_pytorch/encoders/_efficientnet.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"well_formed","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f15a49337e69e937"}},{"code_sha256_prefix":"dbc0ca08d119a5a0","entry":"round_repeats","repo":"qubvel/segmentation_models.pytorch","repo_kind":"listed","path":"segmentation_models_pytorch/encoders/_efficientnet.py","file_url":"https://github.com/qubvel/segmentation_models.pytorch/blob/HEAD/segmentation_models_pytorch/encoders/_efficientnet.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"well_formed","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"dbc0ca08d119a5a0"}},{"code_sha256_prefix":"d892754cd88987af","entry":"adaptive_avgmax_pool2d","repo":"qubvel/segmentation_models.pytorch","repo_kind":"listed","path":"segmentation_models_pytorch/encoders/_dpn.py","file_url":"https://github.com/qubvel/segmentation_models.pytorch/blob/HEAD/segmentation_models_pytorch/encoders/_dpn.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"d892754cd88987af"}},{"code_sha256_prefix":"7366397639253b99","entry":"drop_connect","repo":"qubvel/segmentation_models.pytorch","repo_kind":"listed","path":"segmentation_models_pytorch/encoders/_efficientnet.py","file_url":"https://github.com/qubvel/segmentation_models.pytorch/blob/HEAD/segmentation_models_pytorch/encoders/_efficientnet.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"7366397639253b99"}},{"code_sha256_prefix":"9357c02240b5ef34","entry":"from_pretrained","repo":"qubvel/segmentation_models.pytorch","repo_kind":"listed","path":"segmentation_models_pytorch/base/hub_mixin.py","file_url":"https://github.com/qubvel/segmentation_models.pytorch/blob/HEAD/segmentation_models_pytorch/base/hub_mixin.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"9357c02240b5ef34"}},{"code_sha256_prefix":"1c92dfee49c6912e","entry":"get_norm_layer","repo":"qubvel/segmentation_models.pytorch","repo_kind":"listed","path":"segmentation_models_pytorch/base/modules.py","file_url":"https://github.com/qubvel/segmentation_models.pytorch/blob/HEAD/segmentation_models_pytorch/base/modules.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"1c92dfee49c6912e"}},{"code_sha256_prefix":"3a8857dbf960f1e4","entry":"pooling_factor","repo":"qubvel/segmentation_models.pytorch","repo_kind":"listed","path":"segmentation_models_pytorch/encoders/_dpn.py","file_url":"https://github.com/qubvel/segmentation_models.pytorch/blob/HEAD/segmentation_models_pytorch/encoders/_dpn.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"3a8857dbf960f1e4"}},{"code_sha256_prefix":"13c6bced58eb956a","entry":"supports_config_loading","repo":"qubvel/segmentation_models.pytorch","repo_kind":"listed","path":"segmentation_models_pytorch/base/hub_mixin.py","file_url":"https://github.com/qubvel/segmentation_models.pytorch/blob/HEAD/segmentation_models_pytorch/base/hub_mixin.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"13c6bced58eb956a"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}