{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:2LI53YAVWE4S7L64N3JVZGRI4D","short_pith_number":"pith:2LI53YAV","schema_version":"1.0","canonical_sha256":"d2d1dde015b1392fafdc6ed35c9a28e0dd089d836d5dffb3fa411941985b2738","source":{"kind":"arxiv","id":"2312.01998","version":2},"attestation_state":"computed","paper":{"title":"Language-only Efficient Training of Zero-shot Composed Image Retrieval","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.IR"],"primary_cat":"cs.CV","authors_text":"Geonmo Gu, Sangdoo Yun, Sanghyuk Chun, Wonjae Kim, Yoohoon Kang","submitted_at":"2023-12-04T16:22:06Z","abstract_excerpt":"Composed image retrieval (CIR) task takes a composed query of image and text, aiming to search relative images for both conditions. Conventional CIR approaches need a training dataset composed of triplets of query image, query text, and target image, which is very expensive to collect. Several recent works have worked on the zero-shot (ZS) CIR paradigm to tackle the issue without using pre-collected triplets. However, the existing ZS-CIR methods show limited backbone scalability and generalizability due to the lack of diversity of the input texts during training. We propose a novel CIR framewo"},"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":"2312.01998","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2023-12-04T16:22:06Z","cross_cats_sorted":["cs.IR"],"title_canon_sha256":"137c8556101643c10c90aad3362c27c462500afdaa58e1557fbc42c6ebbe9974","abstract_canon_sha256":"29fcf8bdd9491ebe8f31584389219a28226897e235490ddf6414381aadfb791f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:02:48.871711Z","signature_b64":"SpREDWNnbEakGsekyn/rA1alXG1VeIj8N/HfTYcd2/RzeqoLVP1jjjndMyNF+IkHHSruD/XdRxTqzO/1/oNYBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d2d1dde015b1392fafdc6ed35c9a28e0dd089d836d5dffb3fa411941985b2738","last_reissued_at":"2026-07-05T08:02:48.871242Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:02:48.871242Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Language-only Efficient Training of Zero-shot Composed Image Retrieval","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.IR"],"primary_cat":"cs.CV","authors_text":"Geonmo Gu, Sangdoo Yun, Sanghyuk Chun, Wonjae Kim, Yoohoon Kang","submitted_at":"2023-12-04T16:22:06Z","abstract_excerpt":"Composed image retrieval (CIR) task takes a composed query of image and text, aiming to search relative images for both conditions. Conventional CIR approaches need a training dataset composed of triplets of query image, query text, and target image, which is very expensive to collect. Several recent works have worked on the zero-shot (ZS) CIR paradigm to tackle the issue without using pre-collected triplets. However, the existing ZS-CIR methods show limited backbone scalability and generalizability due to the lack of diversity of the input texts during training. We propose a novel CIR framewo"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.01998","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/2312.01998/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":"2312.01998","created_at":"2026-07-05T08:02:48.871297+00:00"},{"alias_kind":"arxiv_version","alias_value":"2312.01998v2","created_at":"2026-07-05T08:02:48.871297+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.01998","created_at":"2026-07-05T08:02:48.871297+00:00"},{"alias_kind":"pith_short_12","alias_value":"2LI53YAVWE4S","created_at":"2026-07-05T08:02:48.871297+00:00"},{"alias_kind":"pith_short_16","alias_value":"2LI53YAVWE4S7L64","created_at":"2026-07-05T08:02:48.871297+00:00"},{"alias_kind":"pith_short_8","alias_value":"2LI53YAV","created_at":"2026-07-05T08:02:48.871297+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.07032","citing_title":"Never Seen Before: Benchmarking Genuine Zero-Shot Composed Image Retrieval with Consistent Video-Sourced Datasets","ref_index":26,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/2LI53YAVWE4S7L64N3JVZGRI4D","json":"https://pith.science/pith/2LI53YAVWE4S7L64N3JVZGRI4D.json","graph_json":"https://pith.science/api/pith-number/2LI53YAVWE4S7L64N3JVZGRI4D/graph.json","events_json":"https://pith.science/api/pith-number/2LI53YAVWE4S7L64N3JVZGRI4D/events.json","paper":"https://pith.science/paper/2LI53YAV"},"agent_actions":{"view_html":"https://pith.science/pith/2LI53YAVWE4S7L64N3JVZGRI4D","download_json":"https://pith.science/pith/2LI53YAVWE4S7L64N3JVZGRI4D.json","view_paper":"https://pith.science/paper/2LI53YAV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2312.01998&json=true","fetch_graph":"https://pith.science/api/pith-number/2LI53YAVWE4S7L64N3JVZGRI4D/graph.json","fetch_events":"https://pith.science/api/pith-number/2LI53YAVWE4S7L64N3JVZGRI4D/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2LI53YAVWE4S7L64N3JVZGRI4D/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2LI53YAVWE4S7L64N3JVZGRI4D/action/storage_attestation","attest_author":"https://pith.science/pith/2LI53YAVWE4S7L64N3JVZGRI4D/action/author_attestation","sign_citation":"https://pith.science/pith/2LI53YAVWE4S7L64N3JVZGRI4D/action/citation_signature","submit_replication":"https://pith.science/pith/2LI53YAVWE4S7L64N3JVZGRI4D/action/replication_record"}},"created_at":"2026-07-05T08:02:48.871297+00:00","updated_at":"2026-07-05T08:02:48.871297+00:00"}