{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:EPB7O4BXMI65J2SUYSABYCOGJE","short_pith_number":"pith:EPB7O4BX","schema_version":"1.0","canonical_sha256":"23c3f77037623dd4ea54c4801c09c64917b1d051039b2c5dfd18110020a1d161","source":{"kind":"arxiv","id":"2303.02595","version":1},"attestation_state":"computed","paper":{"title":"PyramidFlow: High-Resolution Defect Contrastive Localization using Pyramid Normalizing Flow","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Dong Liu, Jiarui Lei, Xiaobo Hu, Yue Wang","submitted_at":"2023-03-05T07:37:28Z","abstract_excerpt":"During industrial processing, unforeseen defects may arise in products due to uncontrollable factors. Although unsupervised methods have been successful in defect localization, the usual use of pre-trained models results in low-resolution outputs, which damages visual performance. To address this issue, we propose PyramidFlow, the first fully normalizing flow method without pre-trained models that enables high-resolution defect localization. Specifically, we propose a latent template-based defect contrastive localization paradigm to reduce intra-class variance, as the pre-trained models do. In"},"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":"2303.02595","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2023-03-05T07:37:28Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"445d19693dc5b48bbef182a6a201430ad4ee7243d50442da208827fe0a4a657b","abstract_canon_sha256":"b902ceca28be9da0a6f065a3661eeb3e9761602fa4e261ce182f8df81e3fc419"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:48:13.198715Z","signature_b64":"PMMyn1x1ibvaeHbPf2XDDWMooljkUyziYWF3iPYDLd+w97ZIKGuaUHhqleCqReMZB5+KijX0sFgRTrwdZLLuAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"23c3f77037623dd4ea54c4801c09c64917b1d051039b2c5dfd18110020a1d161","last_reissued_at":"2026-07-05T05:48:13.198297Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:48:13.198297Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"PyramidFlow: High-Resolution Defect Contrastive Localization using Pyramid Normalizing Flow","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Dong Liu, Jiarui Lei, Xiaobo Hu, Yue Wang","submitted_at":"2023-03-05T07:37:28Z","abstract_excerpt":"During industrial processing, unforeseen defects may arise in products due to uncontrollable factors. Although unsupervised methods have been successful in defect localization, the usual use of pre-trained models results in low-resolution outputs, which damages visual performance. To address this issue, we propose PyramidFlow, the first fully normalizing flow method without pre-trained models that enables high-resolution defect localization. Specifically, we propose a latent template-based defect contrastive localization paradigm to reduce intra-class variance, as the pre-trained models do. In"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2303.02595","kind":"arxiv","version":1},"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/2303.02595/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":"2303.02595","created_at":"2026-07-05T05:48:13.198362+00:00"},{"alias_kind":"arxiv_version","alias_value":"2303.02595v1","created_at":"2026-07-05T05:48:13.198362+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2303.02595","created_at":"2026-07-05T05:48:13.198362+00:00"},{"alias_kind":"pith_short_12","alias_value":"EPB7O4BXMI65","created_at":"2026-07-05T05:48:13.198362+00:00"},{"alias_kind":"pith_short_16","alias_value":"EPB7O4BXMI65J2SU","created_at":"2026-07-05T05:48:13.198362+00:00"},{"alias_kind":"pith_short_8","alias_value":"EPB7O4BX","created_at":"2026-07-05T05:48:13.198362+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2412.00560","citing_title":"Friend or Foe? Harnessing Controllable Overfitting for Anomaly Detection","ref_index":15,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/EPB7O4BXMI65J2SUYSABYCOGJE","json":"https://pith.science/pith/EPB7O4BXMI65J2SUYSABYCOGJE.json","graph_json":"https://pith.science/api/pith-number/EPB7O4BXMI65J2SUYSABYCOGJE/graph.json","events_json":"https://pith.science/api/pith-number/EPB7O4BXMI65J2SUYSABYCOGJE/events.json","paper":"https://pith.science/paper/EPB7O4BX"},"agent_actions":{"view_html":"https://pith.science/pith/EPB7O4BXMI65J2SUYSABYCOGJE","download_json":"https://pith.science/pith/EPB7O4BXMI65J2SUYSABYCOGJE.json","view_paper":"https://pith.science/paper/EPB7O4BX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2303.02595&json=true","fetch_graph":"https://pith.science/api/pith-number/EPB7O4BXMI65J2SUYSABYCOGJE/graph.json","fetch_events":"https://pith.science/api/pith-number/EPB7O4BXMI65J2SUYSABYCOGJE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EPB7O4BXMI65J2SUYSABYCOGJE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EPB7O4BXMI65J2SUYSABYCOGJE/action/storage_attestation","attest_author":"https://pith.science/pith/EPB7O4BXMI65J2SUYSABYCOGJE/action/author_attestation","sign_citation":"https://pith.science/pith/EPB7O4BXMI65J2SUYSABYCOGJE/action/citation_signature","submit_replication":"https://pith.science/pith/EPB7O4BXMI65J2SUYSABYCOGJE/action/replication_record"}},"created_at":"2026-07-05T05:48:13.198362+00:00","updated_at":"2026-07-05T05:48:13.198362+00:00"}