{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:BUB3AVIFS2LOCCZKUI23VEZ6Q3","short_pith_number":"pith:BUB3AVIF","schema_version":"1.0","canonical_sha256":"0d03b055059696e10b2aa235ba933e86f1f5f2a9d617c24d45a528624c959130","source":{"kind":"arxiv","id":"2103.13859","version":4},"attestation_state":"computed","paper":{"title":"Group-CAM: Group Score-Weighted Visual Explanations for Deep Convolutional Networks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Lu Rao, Qinglong Zhang, Yubin Yang","submitted_at":"2021-03-25T14:16:02Z","abstract_excerpt":"In this paper, we propose an efficient saliency map generation method, called Group score-weighted Class Activation Mapping (Group-CAM), which adopts the \"split-transform-merge\" strategy to generate saliency maps. Specifically, for an input image, the class activations are firstly split into groups. In each group, the sub-activations are summed and de-noised as an initial mask. After that, the initial masks are transformed with meaningful perturbations and then applied to preserve sub-pixels of the input (i.e., masked inputs), which are then fed into the network to calculate the confidence sco"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2103.13859","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2021-03-25T14:16:02Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"5c1f03306e1af6c8c90917e152d7ec8c1a18eb5459e57abfc80739af290c3fc0","abstract_canon_sha256":"09e47c5c8717712e57d42fedccd867ee0af10eaaa2ff32f583162dd9b0f4283f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:50:45.189135Z","signature_b64":"+VR7GaByQMEegc3UQP/b/XFzoKRgJF4c0x26PyZugN+7jGcEsIC1WD/OKhyo4yA0e1B7FcLIbg1kS9iSptBuDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0d03b055059696e10b2aa235ba933e86f1f5f2a9d617c24d45a528624c959130","last_reissued_at":"2026-07-05T02:50:45.188709Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:50:45.188709Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Group-CAM: Group Score-Weighted Visual Explanations for Deep Convolutional Networks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Lu Rao, Qinglong Zhang, Yubin Yang","submitted_at":"2021-03-25T14:16:02Z","abstract_excerpt":"In this paper, we propose an efficient saliency map generation method, called Group score-weighted Class Activation Mapping (Group-CAM), which adopts the \"split-transform-merge\" strategy to generate saliency maps. Specifically, for an input image, the class activations are firstly split into groups. In each group, the sub-activations are summed and de-noised as an initial mask. After that, the initial masks are transformed with meaningful perturbations and then applied to preserve sub-pixels of the input (i.e., masked inputs), which are then fed into the network to calculate the confidence sco"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2103.13859","kind":"arxiv","version":4},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2103.13859/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2103.13859","created_at":"2026-07-05T02:50:45.188774+00:00"},{"alias_kind":"arxiv_version","alias_value":"2103.13859v4","created_at":"2026-07-05T02:50:45.188774+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2103.13859","created_at":"2026-07-05T02:50:45.188774+00:00"},{"alias_kind":"pith_short_12","alias_value":"BUB3AVIFS2LO","created_at":"2026-07-05T02:50:45.188774+00:00"},{"alias_kind":"pith_short_16","alias_value":"BUB3AVIFS2LOCCZK","created_at":"2026-07-05T02:50:45.188774+00:00"},{"alias_kind":"pith_short_8","alias_value":"BUB3AVIF","created_at":"2026-07-05T02:50:45.188774+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.19206","citing_title":"When Can We Trust Deep Neural Networks? Towards Reliable Industrial Deployment with an Interpretability Guide","ref_index":27,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BUB3AVIFS2LOCCZKUI23VEZ6Q3","json":"https://pith.science/pith/BUB3AVIFS2LOCCZKUI23VEZ6Q3.json","graph_json":"https://pith.science/api/pith-number/BUB3AVIFS2LOCCZKUI23VEZ6Q3/graph.json","events_json":"https://pith.science/api/pith-number/BUB3AVIFS2LOCCZKUI23VEZ6Q3/events.json","paper":"https://pith.science/paper/BUB3AVIF"},"agent_actions":{"view_html":"https://pith.science/pith/BUB3AVIFS2LOCCZKUI23VEZ6Q3","download_json":"https://pith.science/pith/BUB3AVIFS2LOCCZKUI23VEZ6Q3.json","view_paper":"https://pith.science/paper/BUB3AVIF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2103.13859&json=true","fetch_graph":"https://pith.science/api/pith-number/BUB3AVIFS2LOCCZKUI23VEZ6Q3/graph.json","fetch_events":"https://pith.science/api/pith-number/BUB3AVIFS2LOCCZKUI23VEZ6Q3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BUB3AVIFS2LOCCZKUI23VEZ6Q3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BUB3AVIFS2LOCCZKUI23VEZ6Q3/action/storage_attestation","attest_author":"https://pith.science/pith/BUB3AVIFS2LOCCZKUI23VEZ6Q3/action/author_attestation","sign_citation":"https://pith.science/pith/BUB3AVIFS2LOCCZKUI23VEZ6Q3/action/citation_signature","submit_replication":"https://pith.science/pith/BUB3AVIFS2LOCCZKUI23VEZ6Q3/action/replication_record"}},"created_at":"2026-07-05T02:50:45.188774+00:00","updated_at":"2026-07-05T02:50:45.188774+00:00"}