{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:BJLOUE3JUAIQZJU64YDVX33L7T","short_pith_number":"pith:BJLOUE3J","schema_version":"1.0","canonical_sha256":"0a56ea1369a0110ca69ee6075bef6bfcccf607a4a6df3503495645e8ebc5378c","source":{"kind":"arxiv","id":"2501.03567","version":2},"attestation_state":"computed","paper":{"title":"Evaluating Image Caption via Cycle-consistent Text-to-Image Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Guo-Hua Wang, Jinbin Bai, Kaifu Zhang, Qing-Guo Chen, Tianyu Cui, Weihua Luo, Ye Shi, Zhao Xu","submitted_at":"2025-01-07T06:35:34Z","abstract_excerpt":"Evaluating image captions typically relies on reference captions, which are costly to obtain and exhibit significant diversity and subjectivity. While reference-free evaluation metrics have been proposed, most focus on cross-modal evaluation between captions and images. Recent research has revealed that the modality gap generally exists in the representation of contrastive learning-based multi-modal systems, undermining the reliability of cross-modality metrics like CLIPScore. In this paper, we propose CAMScore, a cyclic reference-free automatic evaluation metric for image captioning models. 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":"2501.03567","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-01-07T06:35:34Z","cross_cats_sorted":[],"title_canon_sha256":"3456235b169fa40d5caec9b52e556aaceaf4d0d9a79c04092025e10af2d58bb2","abstract_canon_sha256":"73abc52c77f04693e9e95fc2207e025828f539b8ad83a3c7fc0af571d14a8f6c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:58:28.353587Z","signature_b64":"xGGCXQhzIBwGlb23Mqt/dxC8DR65H64nZsg/3YDXHjychOhhS3Y9ZOTic9IDt/1qJ1egE0tUxYJ2+gLd4axmBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0a56ea1369a0110ca69ee6075bef6bfcccf607a4a6df3503495645e8ebc5378c","last_reissued_at":"2026-07-05T09:58:28.353113Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:58:28.353113Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Evaluating Image Caption via Cycle-consistent Text-to-Image Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Guo-Hua Wang, Jinbin Bai, Kaifu Zhang, Qing-Guo Chen, Tianyu Cui, Weihua Luo, Ye Shi, Zhao Xu","submitted_at":"2025-01-07T06:35:34Z","abstract_excerpt":"Evaluating image captions typically relies on reference captions, which are costly to obtain and exhibit significant diversity and subjectivity. While reference-free evaluation metrics have been proposed, most focus on cross-modal evaluation between captions and images. Recent research has revealed that the modality gap generally exists in the representation of contrastive learning-based multi-modal systems, undermining the reliability of cross-modality metrics like CLIPScore. In this paper, we propose CAMScore, a cyclic reference-free automatic evaluation metric for image captioning models. T"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.03567","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/2501.03567/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":"2501.03567","created_at":"2026-07-05T09:58:28.353170+00:00"},{"alias_kind":"arxiv_version","alias_value":"2501.03567v2","created_at":"2026-07-05T09:58:28.353170+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.03567","created_at":"2026-07-05T09:58:28.353170+00:00"},{"alias_kind":"pith_short_12","alias_value":"BJLOUE3JUAIQ","created_at":"2026-07-05T09:58:28.353170+00:00"},{"alias_kind":"pith_short_16","alias_value":"BJLOUE3JUAIQZJU6","created_at":"2026-07-05T09:58:28.353170+00:00"},{"alias_kind":"pith_short_8","alias_value":"BJLOUE3J","created_at":"2026-07-05T09:58:28.353170+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2502.06600","citing_title":"Evaluation of Multilingual Image Captioning: How far can we get with CLIP models?","ref_index":6,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BJLOUE3JUAIQZJU64YDVX33L7T","json":"https://pith.science/pith/BJLOUE3JUAIQZJU64YDVX33L7T.json","graph_json":"https://pith.science/api/pith-number/BJLOUE3JUAIQZJU64YDVX33L7T/graph.json","events_json":"https://pith.science/api/pith-number/BJLOUE3JUAIQZJU64YDVX33L7T/events.json","paper":"https://pith.science/paper/BJLOUE3J"},"agent_actions":{"view_html":"https://pith.science/pith/BJLOUE3JUAIQZJU64YDVX33L7T","download_json":"https://pith.science/pith/BJLOUE3JUAIQZJU64YDVX33L7T.json","view_paper":"https://pith.science/paper/BJLOUE3J","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2501.03567&json=true","fetch_graph":"https://pith.science/api/pith-number/BJLOUE3JUAIQZJU64YDVX33L7T/graph.json","fetch_events":"https://pith.science/api/pith-number/BJLOUE3JUAIQZJU64YDVX33L7T/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BJLOUE3JUAIQZJU64YDVX33L7T/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BJLOUE3JUAIQZJU64YDVX33L7T/action/storage_attestation","attest_author":"https://pith.science/pith/BJLOUE3JUAIQZJU64YDVX33L7T/action/author_attestation","sign_citation":"https://pith.science/pith/BJLOUE3JUAIQZJU64YDVX33L7T/action/citation_signature","submit_replication":"https://pith.science/pith/BJLOUE3JUAIQZJU64YDVX33L7T/action/replication_record"}},"created_at":"2026-07-05T09:58:28.353170+00:00","updated_at":"2026-07-05T09:58:28.353170+00:00"}