{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:R4Z5QKL3LGORSYA67JXIFXT25X","short_pith_number":"pith:R4Z5QKL3","schema_version":"1.0","canonical_sha256":"8f33d8297b599d19601efa6e82de7aedd3d55f2639295f10f255e73db8f7d88f","source":{"kind":"arxiv","id":"2404.06287","version":2},"attestation_state":"computed","paper":{"title":"Counterfactual Reasoning for Multi-Label Image Classification via Patching-Based Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Gang Niu, Jia-Hao Xiao, Masashi Sugiyama, Ming-Kun Xie, Pei Peng, Sheng-Jun Huang","submitted_at":"2024-04-09T13:13:24Z","abstract_excerpt":"The key to multi-label image classification (MLC) is to improve model performance by leveraging label correlations. Unfortunately, it has been shown that overemphasizing co-occurrence relationships can cause the overfitting issue of the model, ultimately leading to performance degradation. In this paper, we provide a causal inference framework to show that the correlative features caused by the target object and its co-occurring objects can be regarded as a mediator, which has both positive and negative impacts on model predictions. On the positive side, the mediator enhances the recognition p"},"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":"2404.06287","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-04-09T13:13:24Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"415014818f30bdca064b5ec369f60fbdd72fcc24fe665ef0774f7bd927650468","abstract_canon_sha256":"0ca7621970a7e67e7dcf217781bd79bf6810a1070dce9f029ce6fd7b05be4b3f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:31:14.394149Z","signature_b64":"B2VY0EAqCk2vb7r1d6ehxzbvoyQJHXUKf48XcreQsckn15FYvAyz2/vvRDxGcDV9Xh/G4to3SMFDirW274PGDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8f33d8297b599d19601efa6e82de7aedd3d55f2639295f10f255e73db8f7d88f","last_reissued_at":"2026-07-05T08:31:14.393664Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:31:14.393664Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Counterfactual Reasoning for Multi-Label Image Classification via Patching-Based Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Gang Niu, Jia-Hao Xiao, Masashi Sugiyama, Ming-Kun Xie, Pei Peng, Sheng-Jun Huang","submitted_at":"2024-04-09T13:13:24Z","abstract_excerpt":"The key to multi-label image classification (MLC) is to improve model performance by leveraging label correlations. Unfortunately, it has been shown that overemphasizing co-occurrence relationships can cause the overfitting issue of the model, ultimately leading to performance degradation. In this paper, we provide a causal inference framework to show that the correlative features caused by the target object and its co-occurring objects can be regarded as a mediator, which has both positive and negative impacts on model predictions. On the positive side, the mediator enhances the recognition p"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.06287","kind":"arxiv","version":2},"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/2404.06287/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":"2404.06287","created_at":"2026-07-05T08:31:14.393724+00:00"},{"alias_kind":"arxiv_version","alias_value":"2404.06287v2","created_at":"2026-07-05T08:31:14.393724+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.06287","created_at":"2026-07-05T08:31:14.393724+00:00"},{"alias_kind":"pith_short_12","alias_value":"R4Z5QKL3LGOR","created_at":"2026-07-05T08:31:14.393724+00:00"},{"alias_kind":"pith_short_16","alias_value":"R4Z5QKL3LGORSYA6","created_at":"2026-07-05T08:31:14.393724+00:00"},{"alias_kind":"pith_short_8","alias_value":"R4Z5QKL3","created_at":"2026-07-05T08:31:14.393724+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.28024","citing_title":"FedHarmony: Harmonizing Heterogeneous Label Correlations in Federated Multi-Label Learning","ref_index":44,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/R4Z5QKL3LGORSYA67JXIFXT25X","json":"https://pith.science/pith/R4Z5QKL3LGORSYA67JXIFXT25X.json","graph_json":"https://pith.science/api/pith-number/R4Z5QKL3LGORSYA67JXIFXT25X/graph.json","events_json":"https://pith.science/api/pith-number/R4Z5QKL3LGORSYA67JXIFXT25X/events.json","paper":"https://pith.science/paper/R4Z5QKL3"},"agent_actions":{"view_html":"https://pith.science/pith/R4Z5QKL3LGORSYA67JXIFXT25X","download_json":"https://pith.science/pith/R4Z5QKL3LGORSYA67JXIFXT25X.json","view_paper":"https://pith.science/paper/R4Z5QKL3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2404.06287&json=true","fetch_graph":"https://pith.science/api/pith-number/R4Z5QKL3LGORSYA67JXIFXT25X/graph.json","fetch_events":"https://pith.science/api/pith-number/R4Z5QKL3LGORSYA67JXIFXT25X/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/R4Z5QKL3LGORSYA67JXIFXT25X/action/timestamp_anchor","attest_storage":"https://pith.science/pith/R4Z5QKL3LGORSYA67JXIFXT25X/action/storage_attestation","attest_author":"https://pith.science/pith/R4Z5QKL3LGORSYA67JXIFXT25X/action/author_attestation","sign_citation":"https://pith.science/pith/R4Z5QKL3LGORSYA67JXIFXT25X/action/citation_signature","submit_replication":"https://pith.science/pith/R4Z5QKL3LGORSYA67JXIFXT25X/action/replication_record"}},"created_at":"2026-07-05T08:31:14.393724+00:00","updated_at":"2026-07-05T08:31:14.393724+00:00"}