{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2018:L2WHSRMYAVGS4U7G3OLRAVPOOA","short_pith_number":"pith:L2WHSRMY","schema_version":"1.0","canonical_sha256":"5eac794598054d2e53e6db971055ee70360330473dc44302a7a543b0ee9ec0d0","source":{"kind":"arxiv","id":"1804.06872","version":3},"attestation_state":"computed","paper":{"title":"Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labels","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Bo Han, Gang Niu, Ivor Tsang, Masashi Sugiyama, Miao Xu, Quanming Yao, Weihua Hu, Xingrui Yu","submitted_at":"2018-04-18T18:42:31Z","abstract_excerpt":"Deep learning with noisy labels is practically challenging, as the capacity of deep models is so high that they can totally memorize these noisy labels sooner or later during training. Nonetheless, recent studies on the memorization effects of deep neural networks show that they would first memorize training data of clean labels and then those of noisy labels. Therefore in this paper, we propose a new deep learning paradigm called Co-teaching for combating with noisy labels. Namely, we train two deep neural networks simultaneously, and let them teach each other given every mini-batch: firstly,"},"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":"1804.06872","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-04-18T18:42:31Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"9fe773781773dcce3c0bf49a927a265f579d8db991e2642ca88d6325f1eb1976","abstract_canon_sha256":"e05b6f318ed5326db9080cae4a3f760edcff7a5ff4240c87a3396e3c5e5fb84b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:02:01.288913Z","signature_b64":"1ujUWbTOuNTDV3YE46BXDgMXLIxHRAfGWXLtsqhz/EucH///KwrPz1jZSZHmjSYtXuqTpM/QSLuFrSzfkps2Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5eac794598054d2e53e6db971055ee70360330473dc44302a7a543b0ee9ec0d0","last_reissued_at":"2026-05-18T00:02:01.288391Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:02:01.288391Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labels","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Bo Han, Gang Niu, Ivor Tsang, Masashi Sugiyama, Miao Xu, Quanming Yao, Weihua Hu, Xingrui Yu","submitted_at":"2018-04-18T18:42:31Z","abstract_excerpt":"Deep learning with noisy labels is practically challenging, as the capacity of deep models is so high that they can totally memorize these noisy labels sooner or later during training. Nonetheless, recent studies on the memorization effects of deep neural networks show that they would first memorize training data of clean labels and then those of noisy labels. Therefore in this paper, we propose a new deep learning paradigm called Co-teaching for combating with noisy labels. Namely, we train two deep neural networks simultaneously, and let them teach each other given every mini-batch: firstly,"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1804.06872","kind":"arxiv","version":3},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"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":"1804.06872","created_at":"2026-05-18T00:02:01.288488+00:00"},{"alias_kind":"arxiv_version","alias_value":"1804.06872v3","created_at":"2026-05-18T00:02:01.288488+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1804.06872","created_at":"2026-05-18T00:02:01.288488+00:00"},{"alias_kind":"pith_short_12","alias_value":"L2WHSRMYAVGS","created_at":"2026-05-18T12:32:33.847187+00:00"},{"alias_kind":"pith_short_16","alias_value":"L2WHSRMYAVGS4U7G","created_at":"2026-05-18T12:32:33.847187+00:00"},{"alias_kind":"pith_short_8","alias_value":"L2WHSRMY","created_at":"2026-05-18T12:32:33.847187+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.25845","citing_title":"Model-agnostic information transfer and fusion for classification with label noise","ref_index":2,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/L2WHSRMYAVGS4U7G3OLRAVPOOA","json":"https://pith.science/pith/L2WHSRMYAVGS4U7G3OLRAVPOOA.json","graph_json":"https://pith.science/api/pith-number/L2WHSRMYAVGS4U7G3OLRAVPOOA/graph.json","events_json":"https://pith.science/api/pith-number/L2WHSRMYAVGS4U7G3OLRAVPOOA/events.json","paper":"https://pith.science/paper/L2WHSRMY"},"agent_actions":{"view_html":"https://pith.science/pith/L2WHSRMYAVGS4U7G3OLRAVPOOA","download_json":"https://pith.science/pith/L2WHSRMYAVGS4U7G3OLRAVPOOA.json","view_paper":"https://pith.science/paper/L2WHSRMY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1804.06872&json=true","fetch_graph":"https://pith.science/api/pith-number/L2WHSRMYAVGS4U7G3OLRAVPOOA/graph.json","fetch_events":"https://pith.science/api/pith-number/L2WHSRMYAVGS4U7G3OLRAVPOOA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/L2WHSRMYAVGS4U7G3OLRAVPOOA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/L2WHSRMYAVGS4U7G3OLRAVPOOA/action/storage_attestation","attest_author":"https://pith.science/pith/L2WHSRMYAVGS4U7G3OLRAVPOOA/action/author_attestation","sign_citation":"https://pith.science/pith/L2WHSRMYAVGS4U7G3OLRAVPOOA/action/citation_signature","submit_replication":"https://pith.science/pith/L2WHSRMYAVGS4U7G3OLRAVPOOA/action/replication_record"}},"created_at":"2026-05-18T00:02:01.288488+00:00","updated_at":"2026-05-18T00:02:01.288488+00:00"}