{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:JCCXV46LLI7V2QNB4QGZUMS2PF","short_pith_number":"pith:JCCXV46L","schema_version":"1.0","canonical_sha256":"48857af3cb5a3f5d41a1e40d9a325a795528a6f5cb3220d563f45a28c96d5362","source":{"kind":"arxiv","id":"2303.17896","version":2},"attestation_state":"computed","paper":{"title":"Exploring the Limits of Deep Image Clustering using Pretrained Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Felix Michels, Hamza Kalisch, Markus Kollmann, Nikolas Adaloglou","submitted_at":"2023-03-31T08:56:29Z","abstract_excerpt":"We present a general methodology that learns to classify images without labels by leveraging pretrained feature extractors. Our approach involves self-distillation training of clustering heads based on the fact that nearest neighbours in the pretrained feature space are likely to share the same label. We propose a novel objective that learns associations between image features by introducing a variant of pointwise mutual information together with instance weighting. We demonstrate that the proposed objective is able to attenuate the effect of false positive pairs while efficiently exploiting t"},"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":"2303.17896","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2023-03-31T08:56:29Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"45c1178ae736f8b5c024acbcf503ecdc9669399b7d2214f84880a70b877c2877","abstract_canon_sha256":"3b175db7cc8f0f32f1ef8f08f0c09a15853ea06d75183b0c61cd89ac8fa92047"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:11:14.466947Z","signature_b64":"bwUurMRIjK+TuTwUR4IybbGNuzrSel9LwVBK9LwiMe/aLhVeo47ebnmPn6lLHqBiGCJBdxiWofQIkrunQRK/AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"48857af3cb5a3f5d41a1e40d9a325a795528a6f5cb3220d563f45a28c96d5362","last_reissued_at":"2026-07-05T07:11:14.466396Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:11:14.466396Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Exploring the Limits of Deep Image Clustering using Pretrained Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Felix Michels, Hamza Kalisch, Markus Kollmann, Nikolas Adaloglou","submitted_at":"2023-03-31T08:56:29Z","abstract_excerpt":"We present a general methodology that learns to classify images without labels by leveraging pretrained feature extractors. Our approach involves self-distillation training of clustering heads based on the fact that nearest neighbours in the pretrained feature space are likely to share the same label. We propose a novel objective that learns associations between image features by introducing a variant of pointwise mutual information together with instance weighting. We demonstrate that the proposed objective is able to attenuate the effect of false positive pairs while efficiently exploiting t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2303.17896","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/2303.17896/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":"2303.17896","created_at":"2026-07-05T07:11:14.466461+00:00"},{"alias_kind":"arxiv_version","alias_value":"2303.17896v2","created_at":"2026-07-05T07:11:14.466461+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2303.17896","created_at":"2026-07-05T07:11:14.466461+00:00"},{"alias_kind":"pith_short_12","alias_value":"JCCXV46LLI7V","created_at":"2026-07-05T07:11:14.466461+00:00"},{"alias_kind":"pith_short_16","alias_value":"JCCXV46LLI7V2QNB","created_at":"2026-07-05T07:11:14.466461+00:00"},{"alias_kind":"pith_short_8","alias_value":"JCCXV46L","created_at":"2026-07-05T07:11:14.466461+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.23198","citing_title":"Label-Efficient Dataset Pruning via Semi-Supervised Pseudo-Labeling","ref_index":31,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JCCXV46LLI7V2QNB4QGZUMS2PF","json":"https://pith.science/pith/JCCXV46LLI7V2QNB4QGZUMS2PF.json","graph_json":"https://pith.science/api/pith-number/JCCXV46LLI7V2QNB4QGZUMS2PF/graph.json","events_json":"https://pith.science/api/pith-number/JCCXV46LLI7V2QNB4QGZUMS2PF/events.json","paper":"https://pith.science/paper/JCCXV46L"},"agent_actions":{"view_html":"https://pith.science/pith/JCCXV46LLI7V2QNB4QGZUMS2PF","download_json":"https://pith.science/pith/JCCXV46LLI7V2QNB4QGZUMS2PF.json","view_paper":"https://pith.science/paper/JCCXV46L","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2303.17896&json=true","fetch_graph":"https://pith.science/api/pith-number/JCCXV46LLI7V2QNB4QGZUMS2PF/graph.json","fetch_events":"https://pith.science/api/pith-number/JCCXV46LLI7V2QNB4QGZUMS2PF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JCCXV46LLI7V2QNB4QGZUMS2PF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JCCXV46LLI7V2QNB4QGZUMS2PF/action/storage_attestation","attest_author":"https://pith.science/pith/JCCXV46LLI7V2QNB4QGZUMS2PF/action/author_attestation","sign_citation":"https://pith.science/pith/JCCXV46LLI7V2QNB4QGZUMS2PF/action/citation_signature","submit_replication":"https://pith.science/pith/JCCXV46LLI7V2QNB4QGZUMS2PF/action/replication_record"}},"created_at":"2026-07-05T07:11:14.466461+00:00","updated_at":"2026-07-05T07:11:14.466461+00:00"}