{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:PVKPIFFDIQPYIBD5AJ32I7SB6U","short_pith_number":"pith:PVKPIFFD","schema_version":"1.0","canonical_sha256":"7d54f414a3441f84047d0277a47e41f50d4b8e1ed76d715336437c49e0d22a86","source":{"kind":"arxiv","id":"2405.19092","version":4},"attestation_state":"computed","paper":{"title":"Benchmarking and Improving Detail Image Caption","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bohong Wu, Haoyuan Guo, Hongyuan Dong, Jiacong Wang, Jiawen Li, Yuan Zhang","submitted_at":"2024-05-29T13:54:12Z","abstract_excerpt":"Image captioning has long been regarded as a fundamental task in visual understanding. Recently, however, few large vision-language model (LVLM) research discusses model's image captioning performance because of the outdated short-caption benchmarks and unreliable evaluation metrics. In this work, we propose to benchmark detail image caption task by curating high-quality evaluation datasets annotated by human experts, GPT-4V and Gemini-1.5-Pro. We also design a more reliable caption evaluation metric called CAPTURE (CAPtion evaluation by exTracting and coUpling coRE information). CAPTURE extra"},"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":"2405.19092","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2024-05-29T13:54:12Z","cross_cats_sorted":[],"title_canon_sha256":"68c1c8f1fcd60d0f297c9ea8e7cbd9737f4bcb501c74e107463badb7fff749d1","abstract_canon_sha256":"6c3c6d509dfef786361b9a73f2cbabcc91eda1c95711b4c382716aff6bdbd4b8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:41:03.353390Z","signature_b64":"bXh4Zdf+SI1YrW7CA3EOBWokPT+YGNGhIzvk4BJE68qRx6rtEVsx3FtmoKOovF6nay+v5MsYDmBx1BKb9QGoCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7d54f414a3441f84047d0277a47e41f50d4b8e1ed76d715336437c49e0d22a86","last_reissued_at":"2026-07-05T08:41:03.352898Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:41:03.352898Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Benchmarking and Improving Detail Image Caption","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bohong Wu, Haoyuan Guo, Hongyuan Dong, Jiacong Wang, Jiawen Li, Yuan Zhang","submitted_at":"2024-05-29T13:54:12Z","abstract_excerpt":"Image captioning has long been regarded as a fundamental task in visual understanding. Recently, however, few large vision-language model (LVLM) research discusses model's image captioning performance because of the outdated short-caption benchmarks and unreliable evaluation metrics. In this work, we propose to benchmark detail image caption task by curating high-quality evaluation datasets annotated by human experts, GPT-4V and Gemini-1.5-Pro. We also design a more reliable caption evaluation metric called CAPTURE (CAPtion evaluation by exTracting and coUpling coRE information). CAPTURE extra"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.19092","kind":"arxiv","version":4},"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/2405.19092/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":"2405.19092","created_at":"2026-07-05T08:41:03.352955+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.19092v4","created_at":"2026-07-05T08:41:03.352955+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.19092","created_at":"2026-07-05T08:41:03.352955+00:00"},{"alias_kind":"pith_short_12","alias_value":"PVKPIFFDIQPY","created_at":"2026-07-05T08:41:03.352955+00:00"},{"alias_kind":"pith_short_16","alias_value":"PVKPIFFDIQPYIBD5","created_at":"2026-07-05T08:41:03.352955+00:00"},{"alias_kind":"pith_short_8","alias_value":"PVKPIFFD","created_at":"2026-07-05T08:41:03.352955+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":16,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.29573","citing_title":"Reliability-Prioritized Fine-Grained Generation in Multimodal Large","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2606.05702","citing_title":"Seeing Time: Benchmarking Chronological Reasoning and Shortcut Biases in Vision-Language Models","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20278","citing_title":"ClaimDiff-RL: Fine-Grained Caption Reinforcement Learning through Visual Claim Comparison","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2605.23694","citing_title":"ChartFI: Benchmarking Faithfulness and Insightfulness of Chart Descriptions from Multimodal Large Language Models","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2606.29573","citing_title":"Reliability-Prioritized Fine-Grained Generation in Multimodal Large","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2605.23694","citing_title":"ChartFI: Benchmarking Faithfulness and Insightfulness of Chart Descriptions from Multimodal Large Language Models","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20278","citing_title":"ClaimDiff-RL: Fine-Grained Caption Reinforcement Learning through Visual Claim Comparison","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2605.14530","citing_title":"Mitigating Mask Prior Drift and Positional Attention Collapse in Large Diffusion Vision-Language Models","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2506.11991","citing_title":"VGR: Visual Grounded Reasoning","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2511.21025","citing_title":"CaptionQA: Is Your Caption as Useful as the Image Itself?","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2605.14530","citing_title":"Mitigating Mask Prior Drift and Positional Attention Collapse in Large Diffusion Vision-Language Models","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13080","citing_title":"Learning to See What You Need: Gaze Attention for Multimodal Large Language Models","ref_index":124,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06080","citing_title":"MSD-Score: Multi-Scale Distributional Scoring for Reference-Free Image Caption Evaluation","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2604.07914","citing_title":"Mitigating Entangled Steering in Large Vision-Language Models for Hallucination Reduction","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2604.05623","citing_title":"DetailVerifyBench: A Benchmark for Dense Hallucination Localization in Long Image Captions","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2604.22855","citing_title":"Evaluating Remote Sensing Image Captions Beyond Metric Biases","ref_index":11,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PVKPIFFDIQPYIBD5AJ32I7SB6U","json":"https://pith.science/pith/PVKPIFFDIQPYIBD5AJ32I7SB6U.json","graph_json":"https://pith.science/api/pith-number/PVKPIFFDIQPYIBD5AJ32I7SB6U/graph.json","events_json":"https://pith.science/api/pith-number/PVKPIFFDIQPYIBD5AJ32I7SB6U/events.json","paper":"https://pith.science/paper/PVKPIFFD"},"agent_actions":{"view_html":"https://pith.science/pith/PVKPIFFDIQPYIBD5AJ32I7SB6U","download_json":"https://pith.science/pith/PVKPIFFDIQPYIBD5AJ32I7SB6U.json","view_paper":"https://pith.science/paper/PVKPIFFD","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.19092&json=true","fetch_graph":"https://pith.science/api/pith-number/PVKPIFFDIQPYIBD5AJ32I7SB6U/graph.json","fetch_events":"https://pith.science/api/pith-number/PVKPIFFDIQPYIBD5AJ32I7SB6U/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PVKPIFFDIQPYIBD5AJ32I7SB6U/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PVKPIFFDIQPYIBD5AJ32I7SB6U/action/storage_attestation","attest_author":"https://pith.science/pith/PVKPIFFDIQPYIBD5AJ32I7SB6U/action/author_attestation","sign_citation":"https://pith.science/pith/PVKPIFFDIQPYIBD5AJ32I7SB6U/action/citation_signature","submit_replication":"https://pith.science/pith/PVKPIFFDIQPYIBD5AJ32I7SB6U/action/replication_record"}},"created_at":"2026-07-05T08:41:03.352955+00:00","updated_at":"2026-07-05T08:41:03.352955+00:00"}