{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:E7PLMPWAY42B64G54H3G2SFDHY","short_pith_number":"pith:E7PLMPWA","schema_version":"1.0","canonical_sha256":"27deb63ec0c7341f70dde1f66d48a33e3b069ef1a1d0bd17ed1d9d4bd4379170","source":{"kind":"arxiv","id":"2207.01426","version":1},"attestation_state":"computed","paper":{"title":"Dynamic Contrastive Distillation for Image-Text Retrieval","license":"http://creativecommons.org/publicdomain/zero/1.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.CV"],"primary_cat":"cs.MM","authors_text":"Dacheng Tao, Jun Rao, Liang Ding, Li Shen, Meng Fang, Shuhan Qi, Yang Liu","submitted_at":"2022-07-04T14:08:59Z","abstract_excerpt":"Although the vision-and-language pretraining (VLP) equipped cross-modal image-text retrieval (ITR) has achieved remarkable progress in the past two years, it suffers from a major drawback: the ever-increasing size of VLP models restricts its deployment to real-world search scenarios (where the high latency is unacceptable). To alleviate this problem, we present a novel plug-in dynamic contrastive distillation (DCD) framework to compress the large VLP models for the ITR task. Technically, we face the following two challenges: 1) the typical uni-modal metric learning approach is difficult to dir"},"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":"2207.01426","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/publicdomain/zero/1.0/","primary_cat":"cs.MM","submitted_at":"2022-07-04T14:08:59Z","cross_cats_sorted":["cs.AI","cs.CL","cs.CV"],"title_canon_sha256":"f4dd784f7753c8c9d8597a5107bf25a9e4c9349de2457e71fc2c04fff32b07a4","abstract_canon_sha256":"7bdbedf10d78eb6e39704489d6c17195e3284170c6055e80ee29580d11214445"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:37:11.784731Z","signature_b64":"rUbqXYMhU7N/nyrkfjquIyMxbJKTC+SBUKiP9xCqWBW/m5bxA9IkZAWz2amxWA/VwX5jzbVAX0gWnXjEwCfaDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"27deb63ec0c7341f70dde1f66d48a33e3b069ef1a1d0bd17ed1d9d4bd4379170","last_reissued_at":"2026-07-05T04:37:11.784324Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:37:11.784324Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Dynamic Contrastive Distillation for Image-Text Retrieval","license":"http://creativecommons.org/publicdomain/zero/1.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.CV"],"primary_cat":"cs.MM","authors_text":"Dacheng Tao, Jun Rao, Liang Ding, Li Shen, Meng Fang, Shuhan Qi, Yang Liu","submitted_at":"2022-07-04T14:08:59Z","abstract_excerpt":"Although the vision-and-language pretraining (VLP) equipped cross-modal image-text retrieval (ITR) has achieved remarkable progress in the past two years, it suffers from a major drawback: the ever-increasing size of VLP models restricts its deployment to real-world search scenarios (where the high latency is unacceptable). To alleviate this problem, we present a novel plug-in dynamic contrastive distillation (DCD) framework to compress the large VLP models for the ITR task. Technically, we face the following two challenges: 1) the typical uni-modal metric learning approach is difficult to dir"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2207.01426","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/2207.01426/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":"2207.01426","created_at":"2026-07-05T04:37:11.784380+00:00"},{"alias_kind":"arxiv_version","alias_value":"2207.01426v1","created_at":"2026-07-05T04:37:11.784380+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2207.01426","created_at":"2026-07-05T04:37:11.784380+00:00"},{"alias_kind":"pith_short_12","alias_value":"E7PLMPWAY42B","created_at":"2026-07-05T04:37:11.784380+00:00"},{"alias_kind":"pith_short_16","alias_value":"E7PLMPWAY42B64G5","created_at":"2026-07-05T04:37:11.784380+00:00"},{"alias_kind":"pith_short_8","alias_value":"E7PLMPWA","created_at":"2026-07-05T04:37:11.784380+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/E7PLMPWAY42B64G54H3G2SFDHY","json":"https://pith.science/pith/E7PLMPWAY42B64G54H3G2SFDHY.json","graph_json":"https://pith.science/api/pith-number/E7PLMPWAY42B64G54H3G2SFDHY/graph.json","events_json":"https://pith.science/api/pith-number/E7PLMPWAY42B64G54H3G2SFDHY/events.json","paper":"https://pith.science/paper/E7PLMPWA"},"agent_actions":{"view_html":"https://pith.science/pith/E7PLMPWAY42B64G54H3G2SFDHY","download_json":"https://pith.science/pith/E7PLMPWAY42B64G54H3G2SFDHY.json","view_paper":"https://pith.science/paper/E7PLMPWA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2207.01426&json=true","fetch_graph":"https://pith.science/api/pith-number/E7PLMPWAY42B64G54H3G2SFDHY/graph.json","fetch_events":"https://pith.science/api/pith-number/E7PLMPWAY42B64G54H3G2SFDHY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/E7PLMPWAY42B64G54H3G2SFDHY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/E7PLMPWAY42B64G54H3G2SFDHY/action/storage_attestation","attest_author":"https://pith.science/pith/E7PLMPWAY42B64G54H3G2SFDHY/action/author_attestation","sign_citation":"https://pith.science/pith/E7PLMPWAY42B64G54H3G2SFDHY/action/citation_signature","submit_replication":"https://pith.science/pith/E7PLMPWAY42B64G54H3G2SFDHY/action/replication_record"}},"created_at":"2026-07-05T04:37:11.784380+00:00","updated_at":"2026-07-05T04:37:11.784380+00:00"}