{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:LQOX6UTFZHQ2ZLOJGKM6TF3PMR","short_pith_number":"pith:LQOX6UTF","schema_version":"1.0","canonical_sha256":"5c1d7f5265c9e1acadc93299e9976f644e0395d30a0677b59838594962b6b8cd","source":{"kind":"arxiv","id":"2312.10032","version":4},"attestation_state":"computed","paper":{"title":"Osprey: Pixel Understanding with Visual Instruction Tuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chi Qin, Dongqi Tang, Jianke Zhu, Jian Liu, Lei Zhang, Wentong Li, Xinjie Luo, Yuqian Yuan","submitted_at":"2023-12-15T18:58:11Z","abstract_excerpt":"Multimodal large language models (MLLMs) have recently achieved impressive general-purpose vision-language capabilities through visual instruction tuning. However, current MLLMs primarily focus on image-level or box-level understanding, falling short in achieving fine-grained vision-language alignment at pixel level. Besides, the lack of mask-based instruction data limits their advancements. In this paper, we propose Osprey, a mask-text instruction tuning approach, to extend MLLMs by incorporating fine-grained mask regions into language instruction, aiming at achieving pixel-wise visual unders"},"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":"2312.10032","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2023-12-15T18:58:11Z","cross_cats_sorted":[],"title_canon_sha256":"260be27a35a27c260cfb95e2fd6023b1753114a1a5cd64ea71ba3458e9b177fd","abstract_canon_sha256":"d53cd9d29fc1789d794dc3bc0cbab6ed8214972e5b91a5d3ac6b9e3138a0a4a4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:05:58.725057Z","signature_b64":"qipm8N1at46ENESpslRkTMSZHzUGE18uat4S1UjbuAVzBGKNl2D6YS7BxRmkotP+GiFPoqHP68uAuVgpWiyyAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5c1d7f5265c9e1acadc93299e9976f644e0395d30a0677b59838594962b6b8cd","last_reissued_at":"2026-07-05T12:05:58.724477Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:05:58.724477Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Osprey: Pixel Understanding with Visual Instruction Tuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chi Qin, Dongqi Tang, Jianke Zhu, Jian Liu, Lei Zhang, Wentong Li, Xinjie Luo, Yuqian Yuan","submitted_at":"2023-12-15T18:58:11Z","abstract_excerpt":"Multimodal large language models (MLLMs) have recently achieved impressive general-purpose vision-language capabilities through visual instruction tuning. However, current MLLMs primarily focus on image-level or box-level understanding, falling short in achieving fine-grained vision-language alignment at pixel level. Besides, the lack of mask-based instruction data limits their advancements. In this paper, we propose Osprey, a mask-text instruction tuning approach, to extend MLLMs by incorporating fine-grained mask regions into language instruction, aiming at achieving pixel-wise visual unders"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.10032","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/2312.10032/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":"2312.10032","created_at":"2026-07-05T12:05:58.724541+00:00"},{"alias_kind":"arxiv_version","alias_value":"2312.10032v4","created_at":"2026-07-05T12:05:58.724541+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.10032","created_at":"2026-07-05T12:05:58.724541+00:00"},{"alias_kind":"pith_short_12","alias_value":"LQOX6UTFZHQ2","created_at":"2026-07-05T12:05:58.724541+00:00"},{"alias_kind":"pith_short_16","alias_value":"LQOX6UTFZHQ2ZLOJ","created_at":"2026-07-05T12:05:58.724541+00:00"},{"alias_kind":"pith_short_8","alias_value":"LQOX6UTF","created_at":"2026-07-05T12:05:58.724541+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2306.13549","citing_title":"A Survey on Multimodal Large Language Models","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2602.22779","citing_title":"TrajTok: Learning Trajectory Tokens enables better Video Understanding","ref_index":82,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LQOX6UTFZHQ2ZLOJGKM6TF3PMR","json":"https://pith.science/pith/LQOX6UTFZHQ2ZLOJGKM6TF3PMR.json","graph_json":"https://pith.science/api/pith-number/LQOX6UTFZHQ2ZLOJGKM6TF3PMR/graph.json","events_json":"https://pith.science/api/pith-number/LQOX6UTFZHQ2ZLOJGKM6TF3PMR/events.json","paper":"https://pith.science/paper/LQOX6UTF"},"agent_actions":{"view_html":"https://pith.science/pith/LQOX6UTFZHQ2ZLOJGKM6TF3PMR","download_json":"https://pith.science/pith/LQOX6UTFZHQ2ZLOJGKM6TF3PMR.json","view_paper":"https://pith.science/paper/LQOX6UTF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2312.10032&json=true","fetch_graph":"https://pith.science/api/pith-number/LQOX6UTFZHQ2ZLOJGKM6TF3PMR/graph.json","fetch_events":"https://pith.science/api/pith-number/LQOX6UTFZHQ2ZLOJGKM6TF3PMR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LQOX6UTFZHQ2ZLOJGKM6TF3PMR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LQOX6UTFZHQ2ZLOJGKM6TF3PMR/action/storage_attestation","attest_author":"https://pith.science/pith/LQOX6UTFZHQ2ZLOJGKM6TF3PMR/action/author_attestation","sign_citation":"https://pith.science/pith/LQOX6UTFZHQ2ZLOJGKM6TF3PMR/action/citation_signature","submit_replication":"https://pith.science/pith/LQOX6UTFZHQ2ZLOJGKM6TF3PMR/action/replication_record"}},"created_at":"2026-07-05T12:05:58.724541+00:00","updated_at":"2026-07-05T12:05:58.724541+00:00"}