{"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/single-image-reflection-removal-exploiting","title":"Single Image Reflection Removal Exploiting Misaligned Training Data and Network Enhancements","arxiv_id":"1904.00637","date":"2019-04-01","proceeding":"CVPR 2019 6","authors":["Kaixuan Wei","Jiaolong Yang","Ying Fu","David Wipf","Hua Huang"],"abstract":"Removing undesirable reflections from a single image captured through a glass\nwindow is of practical importance to visual computing systems. Although\nstate-of-the-art methods can obtain decent results in certain situations,\nperformance declines significantly when tackling more general real-world cases.\nThese failures stem from the intrinsic difficulty of single image reflection\nremoval -- the fundamental ill-posedness of the problem, and the insufficiency\nof densely-labeled training data needed for resolving this ambiguity within\nlearning-based neural network pipelines. In this paper, we address these issues\nby exploiting targeted network enhancements and the novel use of misaligned\ndata. For the former, we augment a baseline network architecture by embedding\ncontext encoding modules that are capable of leveraging high-level contextual\nclues to reduce indeterminacy within areas containing strong reflections. For\nthe latter, we introduce an alignment-invariant loss function that facilitates\nexploiting misaligned real-world training data that is much easier to collect.\nExperimental results collectively show that our method outperforms the\nstate-of-the-art with aligned data, and that significant improvements are\npossible when using additional misaligned data.","url_abs":"http://arxiv.org/abs/1904.00637v1","url_pdf":"http://arxiv.org/pdf/1904.00637v1.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":"single-image-reflection-removal-exploiting","repo_url":"https://github.com/Vandermode/ERRNet","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"reflection-removal","task_name":"Reflection Removal"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/reflection-removal-on-real20","task":"Reflection Removal","dataset":"Real20","model":"ERRNet","rank_in_archive_order":7,"of":8,"metrics":{"PSNR":"22.89","SSIM":"0.803"},"uses_additional_data":false},{"leaderboard":"/sota/reflection-removal-on-sir-2-objects","task":"Reflection Removal","dataset":"SIR^2(Objects)","model":"ERRNet","rank_in_archive_order":5,"of":7,"metrics":{"PSNR":"24.87","SSIM":"0.896"},"uses_additional_data":false},{"leaderboard":"/sota/reflection-removal-on-sir-2-postcard","task":"Reflection Removal","dataset":"SIR^2(Postcard)","model":"ERRNet","rank_in_archive_order":6,"of":6,"metrics":{"PSNR":"22.04","SSIM":"0.876"},"uses_additional_data":false},{"leaderboard":"/sota/reflection-removal-on-sir-2-wild","task":"Reflection Removal","dataset":"SIR^2(Wild)","model":"ERRNet","rank_in_archive_order":6,"of":6,"metrics":{"PSNR":"24.25","SSIM":"0.853"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1904.00637","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1904.00637"}},"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/Vandermode/ERRNet","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran_draft_wrong":1,"ran":2,"unverified":8},"by_repo_kind":{"official":{"samples":11,"ran":3,"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":"91bda7c651d2e3f7","entry":"compute_gradient","repo":"Vandermode/ERRNet","repo_kind":"official","path":"models/losses.py","file_url":"https://github.com/Vandermode/ERRNet/blob/HEAD/models/losses.py","link_basis":"plan_row","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"91bda7c651d2e3f7"}},{"code_sha256_prefix":"805fb9818f986a0d","entry":"receptive_field","repo":"Vandermode/ERRNet","repo_kind":"official","path":"models/networks.py","file_url":"https://github.com/Vandermode/ERRNet/blob/HEAD/models/networks.py","link_basis":"plan_row","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":"805fb9818f986a0d"}},{"code_sha256_prefix":"fde6f58552ebe4f1","entry":"tensor2im","repo":"Vandermode/ERRNet","repo_kind":"official","path":"models/errnet_model.py","file_url":"https://github.com/Vandermode/ERRNet/blob/HEAD/models/errnet_model.py","link_basis":"plan_row","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":"fde6f58552ebe4f1"}},{"code_sha256_prefix":"573d768198fda403","entry":"CX_loss","repo":"Vandermode/ERRNet","repo_kind":"official","path":"models/CX/CX_distance.py","file_url":"https://github.com/Vandermode/ERRNet/blob/HEAD/models/CX/CX_distance.py","link_basis":"first_harvest_node","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":"573d768198fda403"}},{"code_sha256_prefix":"7a92eb4c93e6b804","entry":"crop_quarters","repo":"Vandermode/ERRNet","repo_kind":"official","path":"models/CX/CX_helper.py","file_url":"https://github.com/Vandermode/ERRNet/blob/HEAD/models/CX/CX_helper.py","link_basis":"first_harvest_node","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":"7a92eb4c93e6b804"}},{"code_sha256_prefix":"2c137d3584064330","entry":"define_D","repo":"Vandermode/ERRNet","repo_kind":"official","path":"models/networks.py","file_url":"https://github.com/Vandermode/ERRNet/blob/HEAD/models/networks.py","link_basis":"first_harvest_node","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":"2c137d3584064330"}},{"code_sha256_prefix":"7702313c16b1182f","entry":"get_norm_layer","repo":"Vandermode/ERRNet","repo_kind":"official","path":"models/networks.py","file_url":"https://github.com/Vandermode/ERRNet/blob/HEAD/models/networks.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":"7702313c16b1182f"}},{"code_sha256_prefix":"310ba2102b406358","entry":"init_loss","repo":"Vandermode/ERRNet","repo_kind":"official","path":"models/losses.py","file_url":"https://github.com/Vandermode/ERRNet/blob/HEAD/models/losses.py","link_basis":"first_harvest_node","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":"310ba2102b406358"}},{"code_sha256_prefix":"07f5bda268e05ef2","entry":"random_pooling","repo":"Vandermode/ERRNet","repo_kind":"official","path":"models/CX/CX_helper.py","file_url":"https://github.com/Vandermode/ERRNet/blob/HEAD/models/CX/CX_helper.py","link_basis":"first_harvest_node","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":"07f5bda268e05ef2"}},{"code_sha256_prefix":"5b0d0624510bbe25","entry":"random_sampling","repo":"Vandermode/ERRNet","repo_kind":"official","path":"models/CX/CX_helper.py","file_url":"https://github.com/Vandermode/ERRNet/blob/HEAD/models/CX/CX_helper.py","link_basis":"first_harvest_node","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":"5b0d0624510bbe25"}},{"code_sha256_prefix":"33442b837e557c99","entry":"symetric_CX_loss","repo":"Vandermode/ERRNet","repo_kind":"official","path":"models/CX/CX_distance.py","file_url":"https://github.com/Vandermode/ERRNet/blob/HEAD/models/CX/CX_distance.py","link_basis":"first_harvest_node","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":"33442b837e557c99"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}