{"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/less-is-more-fewer-interpretable-region-via","title":"Less is More: Fewer Interpretable Region via Submodular Subset Selection","arxiv_id":"2402.09164","date":"2024-02-14","proceeding":null,"authors":["Ruoyu Chen","Hua Zhang","Siyuan Liang","Jingzhi Li","Xiaochun Cao"],"abstract":"Image attribution algorithms aim to identify important regions that are highly relevant to model decisions. Although existing attribution solutions can effectively assign importance to target elements, they still face the following challenges: 1) existing attribution methods generate inaccurate small regions thus misleading the direction of correct attribution, and 2) the model cannot produce good attribution results for samples with wrong predictions. To address the above challenges, this paper re-models the above image attribution problem as a submodular subset selection problem, aiming to enhance model interpretability using fewer regions. To address the lack of attention to local regions, we construct a novel submodular function to discover more accurate small interpretation regions. To enhance the attribution effect for all samples, we also impose four different constraints on the selection of sub-regions, i.e., confidence, effectiveness, consistency, and collaboration scores, to assess the importance of various subsets. Moreover, our theoretical analysis substantiates that the proposed function is in fact submodular. Extensive experiments show that the proposed method outperforms SOTA methods on two face datasets (Celeb-A and VGG-Face2) and one fine-grained dataset (CUB-200-2011). For correctly predicted samples, the proposed method improves the Deletion and Insertion scores with an average of 4.9% and 2.5% gain relative to HSIC-Attribution. For incorrectly predicted samples, our method achieves gains of 81.0% and 18.4% compared to the HSIC-Attribution algorithm in the average highest confidence and Insertion score respectively. The code is released at https://github.com/RuoyuChen10/SMDL-Attribution.","url_abs":"https://arxiv.org/abs/2402.09164v3","url_pdf":"https://arxiv.org/pdf/2402.09164v3.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":"less-is-more-fewer-interpretable-region-via","repo_url":"https://github.com/ruoyuchen10/smdl-attribution","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"error-understanding","task_name":"Error Understanding"},{"task_slug":"image-attribution","task_name":"Image Attribution"},{"task_slug":"interpretability-techniques-for-deep-learning","task_name":"Interpretability Techniques for Deep Learning"}],"methods":[{"method_slug":"cam","method_name":"CAM"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/error-understanding-on-cub-200-2011-1","task":"Error Understanding","dataset":"CUB-200-2011","model":"SMDL-Attribution (ICLR version)","rank_in_archive_order":1,"of":4,"metrics":{"Average highest confidence (EfficientNetV2-M)":"0.3306","Average highest confidence (MobileNetV2)":"0.5367","Average highest confidence (ResNet-101)":"0.4513","Insertion AUC score (EfficientNetV2-M)":"0.1748","Insertion AUC score (MobileNetV2)":"0.1922","Insertion AUC score (ResNet-101)":"0.1772"},"uses_additional_data":false},{"leaderboard":"/sota/image-attribution-on-cub-200-2011-1","task":"Image Attribution","dataset":"CUB-200-2011","model":"SMDL-Attribution (ICLR version)","rank_in_archive_order":1,"of":8,"metrics":{"Deletion AUC score (ResNet-101)":"0.0613","Insertion AUC score (ResNet-101)":"0.7262"},"uses_additional_data":false},{"leaderboard":"/sota/image-attribution-on-celeba","task":"Image Attribution","dataset":"CelebA","model":"SMDL-Attribution (ICLR version)","rank_in_archive_order":1,"of":8,"metrics":{"Deletion AUC score (ArcFace ResNet-101)":"0.1054","Insertion AUC score (ArcFace ResNet-101)":"0.5752"},"uses_additional_data":false},{"leaderboard":"/sota/image-attribution-on-vggface2","task":"Image Attribution","dataset":"VGGFace2","model":"SMDL-Attribution (ICLR version)","rank_in_archive_order":1,"of":8,"metrics":{"Deletion AUC score (ArcFace ResNet-101)":"0.1304","Insertion AUC score (ArcFace ResNet-101)":"0.6705"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/2402.09164","atlas_url":"https://app.syntology.ai/?focus=2402.09164","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2402.09164"}},"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":"deterministic:regex_extraction","url":"https://github.com/RuoyuChen10/SMDL-Attribution","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/ruoyuchen10/smdl-attribution","reach":null}],"summary":{"ran":2,"ran_draft_wrong":1},"by_repo_kind":{"official":{"samples":3,"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":3,"samples":[{"code_sha256_prefix":"da8697bef999a4ea","entry":"MultiModalSubModularExplanation","repo":"ruoyuchen10/smdl-attribution","repo_kind":"official","path":"models/submodular_vit_efficient.py","file_url":"https://github.com/ruoyuchen10/smdl-attribution/blob/HEAD/models/submodular_vit_efficient.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"da8697bef999a4ea"}},{"code_sha256_prefix":"11a6722521b17c5e","entry":"MultiModalSubModularExplanationEfficientV1","repo":"ruoyuchen10/smdl-attribution","repo_kind":"official","path":"models/submodular_vit_efficient.py","file_url":"https://github.com/ruoyuchen10/smdl-attribution/blob/HEAD/models/submodular_vit_efficient.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"11a6722521b17c5e"}},{"code_sha256_prefix":"4e71c97a79591777","entry":"Partition_by_patch","repo":"RuoyuChen10/SMDL-Attribution","repo_kind":"official","path":"submodular_attribution/smdl_explanation_celeba.py","file_url":"https://github.com/RuoyuChen10/SMDL-Attribution/blob/HEAD/submodular_attribution/smdl_explanation_celeba.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"4e71c97a79591777"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}