{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:AOD3MAP6D62QG6D6EPTR6YNQY2","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"e88a9903116d6c10a4f38a8cb17156b53f7609b5805a39ad19743508dbf4a4f7","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2023-03-23T17:59:53Z","title_canon_sha256":"3d859544338402297a7f9a88ef123ec11903d1ef946c626107b37d314c720ca8"},"schema_version":"1.0","source":{"id":"2303.13518","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2303.13518","created_at":"2026-07-05T05:54:03Z"},{"alias_kind":"arxiv_version","alias_value":"2303.13518v1","created_at":"2026-07-05T05:54:03Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2303.13518","created_at":"2026-07-05T05:54:03Z"},{"alias_kind":"pith_short_12","alias_value":"AOD3MAP6D62Q","created_at":"2026-07-05T05:54:03Z"},{"alias_kind":"pith_short_16","alias_value":"AOD3MAP6D62QG6D6","created_at":"2026-07-05T05:54:03Z"},{"alias_kind":"pith_short_8","alias_value":"AOD3MAP6","created_at":"2026-07-05T05:54:03Z"}],"graph_snapshots":[{"event_id":"sha256:aa5a308ddef83229cb3d75061eb8dfab1e17a2fab260bd7e78149e08961bdf40","target":"graph","created_at":"2026-07-05T05:54:03Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2303.13518/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The core problem in zero-shot open vocabulary detection is how to align visual and text features, so that the detector performs well on unseen classes. Previous approaches train the feature pyramid and detection head from scratch, which breaks the vision-text feature alignment established during pretraining, and struggles to prevent the language model from forgetting unseen classes.\n  We propose three methods to alleviate these issues. Firstly, a simple scheme is used to augment the text embeddings which prevents overfitting to a small number of classes seen during training, while simultaneous","authors_text":"Alex Andonian, Andrew Zisserman, Arthur Mensch, Jean-Baptiste Alayrac, Olivier J. H\\'enaff, Relja Arandjelovi\\'c","cross_cats":["cs.AI","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2023-03-23T17:59:53Z","title":"Three ways to improve feature alignment for open vocabulary detection"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2303.13518","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:9754355a22427d496ae11cb56844e6ca4653970ea8825d73ecbb8b46c7424ea4","target":"record","created_at":"2026-07-05T05:54:03Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"e88a9903116d6c10a4f38a8cb17156b53f7609b5805a39ad19743508dbf4a4f7","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2023-03-23T17:59:53Z","title_canon_sha256":"3d859544338402297a7f9a88ef123ec11903d1ef946c626107b37d314c720ca8"},"schema_version":"1.0","source":{"id":"2303.13518","kind":"arxiv","version":1}},"canonical_sha256":"0387b601fe1fb503787e23e71f61b0c6a862f7c362d6339b8ac96615e23659ae","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"0387b601fe1fb503787e23e71f61b0c6a862f7c362d6339b8ac96615e23659ae","first_computed_at":"2026-07-05T05:54:03.245821Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T05:54:03.245821Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"aGOIqPp/7veJKAgBtDt5BxSRvXuQWTOae9C99TJ9dHLD4pMNjJGDwB0+qTD7EMC2EJE6IZiB81jQjFsp3KgeAA==","signature_status":"signed_v1","signed_at":"2026-07-05T05:54:03.246307Z","signed_message":"canonical_sha256_bytes"},"source_id":"2303.13518","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:9754355a22427d496ae11cb56844e6ca4653970ea8825d73ecbb8b46c7424ea4","sha256:aa5a308ddef83229cb3d75061eb8dfab1e17a2fab260bd7e78149e08961bdf40"],"state_sha256":"3f5380e3470a780f20880deaf2b95b505212e65d733e0b57e787da692bf4cabc"}