{"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/humus-net-hybrid-unrolled-multi-scale-network","title":"HUMUS-Net: Hybrid unrolled multi-scale network architecture for accelerated MRI reconstruction","arxiv_id":"2203.08213","date":"2022-03-15","proceeding":null,"authors":["Zalan Fabian","Berk Tınaz","Mahdi Soltanolkotabi"],"abstract":"In accelerated MRI reconstruction, the anatomy of a patient is recovered from a set of under-sampled and noisy measurements. Deep learning approaches have been proven to be successful in solving this ill-posed inverse problem and are capable of producing very high quality reconstructions. However, current architectures heavily rely on convolutions, that are content-independent and have difficulties modeling long-range dependencies in images. Recently, Transformers, the workhorse of contemporary natural language processing, have emerged as powerful building blocks for a multitude of vision tasks. These models split input images into non-overlapping patches, embed the patches into lower-dimensional tokens and utilize a self-attention mechanism that does not suffer from the aforementioned weaknesses of convolutional architectures. However, Transformers incur extremely high compute and memory cost when 1) the input image resolution is high and 2) when the image needs to be split into a large number of patches to preserve fine detail information, both of which are typical in low-level vision problems such as MRI reconstruction, having a compounding effect. To tackle these challenges, we propose HUMUS-Net, a hybrid architecture that combines the beneficial implicit bias and efficiency of convolutions with the power of Transformer blocks in an unrolled and multi-scale network. HUMUS-Net extracts high-resolution features via convolutional blocks and refines low-resolution features via a novel Transformer-based multi-scale feature extractor. Features from both levels are then synthesized into a high-resolution output reconstruction. Our network establishes new state of the art on the largest publicly available MRI dataset, the fastMRI dataset. We further demonstrate the performance of HUMUS-Net on two other popular MRI datasets and perform fine-grained ablation studies to validate our design.","url_abs":"https://arxiv.org/abs/2203.08213v2","url_pdf":"https://arxiv.org/pdf/2203.08213v2.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":"humus-net-hybrid-unrolled-multi-scale-network","repo_url":"https://github.com/MathFLDS/HUMUS-Net","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"humus-net-hybrid-unrolled-multi-scale-network","repo_url":"https://github.com/z-fabian/HUMUS-Net","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"anatomy","task_name":"Anatomy"},{"task_slug":"mri-reconstruction","task_name":"MRI Reconstruction"}],"methods":[{"method_slug":"absolute-position-encodings","method_name":"Absolute Position Encodings"},{"method_slug":"adam","method_name":"Adam"},{"method_slug":"attention","method_name":"Attention"},{"method_slug":"bpe","method_name":"BPE"},{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"dropout","method_name":"Dropout"},{"method_slug":"label-smoothing","method_name":"Label Smoothing"},{"method_slug":"layer-normalization","method_name":"Layer Normalization"},{"method_slug":"linear-layer","method_name":"Linear Layer"},{"method_slug":"multi-head-attention","method_name":"Multi-Head Attention"},{"method_slug":"position-wise-feed-forward-layer","method_name":"Position-Wise Feed-Forward Layer"},{"method_slug":"residual-connection","method_name":"Residual Connection"},{"method_slug":"softmax","method_name":"Softmax"},{"method_slug":"transformer","method_name":"Transformer"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/mri-reconstruction-on-fastmri-knee-8x","task":"MRI Reconstruction","dataset":"fastMRI Knee 8x","model":"HUMUS-Net (train+val data)","rank_in_archive_order":1,"of":4,"metrics":{"PSNR":"37.3","SSIM":"0.8945"},"uses_additional_data":true},{"leaderboard":"/sota/mri-reconstruction-on-fastmri-knee-8x","task":"MRI Reconstruction","dataset":"fastMRI Knee 8x","model":"HUMUS-Net (train only)","rank_in_archive_order":2,"of":4,"metrics":{"PSNR":"37.0","SSIM":"0.8936"},"uses_additional_data":false},{"leaderboard":"/sota/mri-reconstruction-on-fastmri-knee-val-8x","task":"MRI Reconstruction","dataset":"fastMRI Knee Val 8x","model":"HUMUS-Net-L","rank_in_archive_order":2,"of":4,"metrics":{"NMSE":"0.0086","PSNR":"37.45","Params (M)":"228","SSIM":"0.8955"},"uses_additional_data":false},{"leaderboard":"/sota/mri-reconstruction-on-fastmri-knee-val-8x","task":"MRI Reconstruction","dataset":"fastMRI Knee Val 8x","model":"HUMUS-Net","rank_in_archive_order":3,"of":4,"metrics":{"NMSE":"0.0090","PSNR":"37.20","Params (M)":"109","SSIM":"0.8946"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/2203.08213","atlas_url":"https://app.syntology.ai/?focus=2203.08213","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2203.08213"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+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/MathFLDS/HUMUS-Net","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/z-fabian/HUMUS-Net","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran":3,"ran_fixture":1,"ran_draft_wrong":1},"by_repo_kind":{"official":{"samples":5,"ran":5,"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":"f8ec18034e33feb4","entry":"batched_mask_center","repo":"MathFLDS/HUMUS-Net","repo_kind":"official","path":"models/humus_net.py","file_url":"https://github.com/MathFLDS/HUMUS-Net/blob/HEAD/models/humus_net.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f8ec18034e33feb4"}},{"code_sha256_prefix":"1b315f2b67a89eab","entry":"load_args_from_config","repo":"MathFLDS/HUMUS-Net","repo_kind":"official","path":"humus_examples/utils.py","file_url":"https://github.com/MathFLDS/HUMUS-Net/blob/HEAD/humus_examples/utils.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"1b315f2b67a89eab"}},{"code_sha256_prefix":"bdf81e063710a085","entry":"mask_center","repo":"MathFLDS/HUMUS-Net","repo_kind":"official","path":"models/humus_net.py","file_url":"https://github.com/MathFLDS/HUMUS-Net/blob/HEAD/models/humus_net.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"bdf81e063710a085"}},{"code_sha256_prefix":"f9fd6241d935f07b","entry":"window_partition","repo":"MathFLDS/HUMUS-Net","repo_kind":"official","path":"models/humus_block.py","file_url":"https://github.com/MathFLDS/HUMUS-Net/blob/HEAD/models/humus_block.py","link_basis":"harvester_set","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f9fd6241d935f07b"}},{"code_sha256_prefix":"fb32094c6dbece71","entry":"window_reverse","repo":"MathFLDS/HUMUS-Net","repo_kind":"official","path":"models/humus_block.py","file_url":"https://github.com/MathFLDS/HUMUS-Net/blob/HEAD/models/humus_block.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"invariant","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"fb32094c6dbece71"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}