{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:RALQY4OCSY4OTEYMLIG6M7MDJ4","short_pith_number":"pith:RALQY4OC","schema_version":"1.0","canonical_sha256":"88170c71c29638e9930c5a0de67d834f317dc77ea0ffe7278b95fe68e8515f7a","source":{"kind":"arxiv","id":"2502.06445","version":1},"attestation_state":"computed","paper":{"title":"Benchmarking Vision-Language Models on Optical Character Recognition in Dynamic Video Environments","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Ashish Choithani, Ashutosh Trivedi, Augustya Sharma, Sankalp Nagaonkar","submitted_at":"2025-02-10T13:20:19Z","abstract_excerpt":"This paper introduces an open-source benchmark for evaluating Vision-Language Models (VLMs) on Optical Character Recognition (OCR) tasks in dynamic video environments. We present a curated dataset containing 1,477 manually annotated frames spanning diverse domains, including code editors, news broadcasts, YouTube videos, and advertisements. Three state of the art VLMs - Claude-3, Gemini-1.5, and GPT-4o are benchmarked against traditional OCR systems such as EasyOCR and RapidOCR. Evaluation metrics include Word Error Rate (WER), Character Error Rate (CER), and Accuracy. Our results highlight th"},"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":"2502.06445","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-02-10T13:20:19Z","cross_cats_sorted":[],"title_canon_sha256":"61d04123c16567882afd0a85a1e6d507f172928e1b1c177fcfd6baed97d7527f","abstract_canon_sha256":"26d0e97b4f916fc439c097865a536309f2f73a7d21bf50be4f36ce88f8fd9ffa"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:12:06.762773Z","signature_b64":"1kg3b2dcExCTqBrVCzuKYLEtP0imuF1KgbrO5aYhn1VodfSUw01TLYgtk5Y5+X+HrFYVP2K13v22cJBELstsCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"88170c71c29638e9930c5a0de67d834f317dc77ea0ffe7278b95fe68e8515f7a","last_reissued_at":"2026-07-05T10:12:06.762283Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:12:06.762283Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Benchmarking Vision-Language Models on Optical Character Recognition in Dynamic Video Environments","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Ashish Choithani, Ashutosh Trivedi, Augustya Sharma, Sankalp Nagaonkar","submitted_at":"2025-02-10T13:20:19Z","abstract_excerpt":"This paper introduces an open-source benchmark for evaluating Vision-Language Models (VLMs) on Optical Character Recognition (OCR) tasks in dynamic video environments. We present a curated dataset containing 1,477 manually annotated frames spanning diverse domains, including code editors, news broadcasts, YouTube videos, and advertisements. Three state of the art VLMs - Claude-3, Gemini-1.5, and GPT-4o are benchmarked against traditional OCR systems such as EasyOCR and RapidOCR. Evaluation metrics include Word Error Rate (WER), Character Error Rate (CER), and Accuracy. Our results highlight th"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.06445","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/2502.06445/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":"2502.06445","created_at":"2026-07-05T10:12:06.762344+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.06445v1","created_at":"2026-07-05T10:12:06.762344+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.06445","created_at":"2026-07-05T10:12:06.762344+00:00"},{"alias_kind":"pith_short_12","alias_value":"RALQY4OCSY4O","created_at":"2026-07-05T10:12:06.762344+00:00"},{"alias_kind":"pith_short_16","alias_value":"RALQY4OCSY4OTEYM","created_at":"2026-07-05T10:12:06.762344+00:00"},{"alias_kind":"pith_short_8","alias_value":"RALQY4OC","created_at":"2026-07-05T10:12:06.762344+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.19844","citing_title":"If you're waiting for a sign... that might not be it! Mitigating Trust Boundary Confusion from Visual Injections on Vision-Language Agentic Systems","ref_index":33,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/RALQY4OCSY4OTEYMLIG6M7MDJ4","json":"https://pith.science/pith/RALQY4OCSY4OTEYMLIG6M7MDJ4.json","graph_json":"https://pith.science/api/pith-number/RALQY4OCSY4OTEYMLIG6M7MDJ4/graph.json","events_json":"https://pith.science/api/pith-number/RALQY4OCSY4OTEYMLIG6M7MDJ4/events.json","paper":"https://pith.science/paper/RALQY4OC"},"agent_actions":{"view_html":"https://pith.science/pith/RALQY4OCSY4OTEYMLIG6M7MDJ4","download_json":"https://pith.science/pith/RALQY4OCSY4OTEYMLIG6M7MDJ4.json","view_paper":"https://pith.science/paper/RALQY4OC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.06445&json=true","fetch_graph":"https://pith.science/api/pith-number/RALQY4OCSY4OTEYMLIG6M7MDJ4/graph.json","fetch_events":"https://pith.science/api/pith-number/RALQY4OCSY4OTEYMLIG6M7MDJ4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RALQY4OCSY4OTEYMLIG6M7MDJ4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RALQY4OCSY4OTEYMLIG6M7MDJ4/action/storage_attestation","attest_author":"https://pith.science/pith/RALQY4OCSY4OTEYMLIG6M7MDJ4/action/author_attestation","sign_citation":"https://pith.science/pith/RALQY4OCSY4OTEYMLIG6M7MDJ4/action/citation_signature","submit_replication":"https://pith.science/pith/RALQY4OCSY4OTEYMLIG6M7MDJ4/action/replication_record"}},"created_at":"2026-07-05T10:12:06.762344+00:00","updated_at":"2026-07-05T10:12:06.762344+00:00"}