{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:2KSGYOPGQ3K74R2UCHAOTV57ZB","short_pith_number":"pith:2KSGYOPG","schema_version":"1.0","canonical_sha256":"d2a46c39e686d5fe475411c0e9d7bfc8533a0c1a384ef59e0cb27a16331a2826","source":{"kind":"arxiv","id":"2407.04538","version":3},"attestation_state":"computed","paper":{"title":"PDiscoFormer: Relaxing Part Discovery Constraints with Vision Transformers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Ananthu Aniraj, Cassio F.Dantas, Diego Marcos, Dino Ienco","submitted_at":"2024-07-05T14:24:37Z","abstract_excerpt":"Computer vision methods that explicitly detect object parts and reason on them are a step towards inherently interpretable models. Existing approaches that perform part discovery driven by a fine-grained classification task make very restrictive assumptions on the geometric properties of the discovered parts; they should be small and compact. Although this prior is useful in some cases, in this paper we show that pre-trained transformer-based vision models, such as self-supervised DINOv2 ViT, enable the relaxation of these constraints. In particular, we find that a total variation (TV) prior, "},"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":"2407.04538","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-07-05T14:24:37Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"4e3ea06bea4bc1b18d0d376a0a49f9117a18941bbe2a2fa1740e128cd6bea120","abstract_canon_sha256":"74d00ebb4f6e07827241ad4d4c6b370ab8fa9e5c94ae8eba477ffb2205aaa453"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:46:55.909970Z","signature_b64":"NeJkGRhLfT3ilQWzcPhmx4X7yxdCF1q1Jd331te+oa8k+3izDEVW19hmQ6UL9S71qLLSaC/FmnP6ZOi3aGYNAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d2a46c39e686d5fe475411c0e9d7bfc8533a0c1a384ef59e0cb27a16331a2826","last_reissued_at":"2026-07-05T08:46:55.909496Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:46:55.909496Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"PDiscoFormer: Relaxing Part Discovery Constraints with Vision Transformers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Ananthu Aniraj, Cassio F.Dantas, Diego Marcos, Dino Ienco","submitted_at":"2024-07-05T14:24:37Z","abstract_excerpt":"Computer vision methods that explicitly detect object parts and reason on them are a step towards inherently interpretable models. Existing approaches that perform part discovery driven by a fine-grained classification task make very restrictive assumptions on the geometric properties of the discovered parts; they should be small and compact. Although this prior is useful in some cases, in this paper we show that pre-trained transformer-based vision models, such as self-supervised DINOv2 ViT, enable the relaxation of these constraints. In particular, we find that a total variation (TV) prior, "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.04538","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":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2407.04538/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":"2407.04538","created_at":"2026-07-05T08:46:55.909563+00:00"},{"alias_kind":"arxiv_version","alias_value":"2407.04538v3","created_at":"2026-07-05T08:46:55.909563+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.04538","created_at":"2026-07-05T08:46:55.909563+00:00"},{"alias_kind":"pith_short_12","alias_value":"2KSGYOPGQ3K7","created_at":"2026-07-05T08:46:55.909563+00:00"},{"alias_kind":"pith_short_16","alias_value":"2KSGYOPGQ3K74R2U","created_at":"2026-07-05T08:46:55.909563+00:00"},{"alias_kind":"pith_short_8","alias_value":"2KSGYOPG","created_at":"2026-07-05T08:46:55.909563+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2507.06928","citing_title":"Adaptive Part Learning for Fine-Grained Generalized Category Discovery: A Plug-and-Play Enhancement","ref_index":1,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/2KSGYOPGQ3K74R2UCHAOTV57ZB","json":"https://pith.science/pith/2KSGYOPGQ3K74R2UCHAOTV57ZB.json","graph_json":"https://pith.science/api/pith-number/2KSGYOPGQ3K74R2UCHAOTV57ZB/graph.json","events_json":"https://pith.science/api/pith-number/2KSGYOPGQ3K74R2UCHAOTV57ZB/events.json","paper":"https://pith.science/paper/2KSGYOPG"},"agent_actions":{"view_html":"https://pith.science/pith/2KSGYOPGQ3K74R2UCHAOTV57ZB","download_json":"https://pith.science/pith/2KSGYOPGQ3K74R2UCHAOTV57ZB.json","view_paper":"https://pith.science/paper/2KSGYOPG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2407.04538&json=true","fetch_graph":"https://pith.science/api/pith-number/2KSGYOPGQ3K74R2UCHAOTV57ZB/graph.json","fetch_events":"https://pith.science/api/pith-number/2KSGYOPGQ3K74R2UCHAOTV57ZB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2KSGYOPGQ3K74R2UCHAOTV57ZB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2KSGYOPGQ3K74R2UCHAOTV57ZB/action/storage_attestation","attest_author":"https://pith.science/pith/2KSGYOPGQ3K74R2UCHAOTV57ZB/action/author_attestation","sign_citation":"https://pith.science/pith/2KSGYOPGQ3K74R2UCHAOTV57ZB/action/citation_signature","submit_replication":"https://pith.science/pith/2KSGYOPGQ3K74R2UCHAOTV57ZB/action/replication_record"}},"created_at":"2026-07-05T08:46:55.909563+00:00","updated_at":"2026-07-05T08:46:55.909563+00:00"}