{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:OB2YA6VL2AK363O3NEYCPBV7OK","short_pith_number":"pith:OB2YA6VL","schema_version":"1.0","canonical_sha256":"7075807aabd015bf6ddb69302786bf72bfaa2a62d36ce1c8b94fae551545b1e0","source":{"kind":"arxiv","id":"2506.03589","version":3},"attestation_state":"computed","paper":{"title":"BiMa: Towards Biases Mitigation for Text-Video Retrieval via Scene Element Guidance","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.CV","authors_text":"Anh Nguyen, Huy Le, Ngan Le, Nhat Chung, Tung Kieu","submitted_at":"2025-06-04T05:40:54Z","abstract_excerpt":"Text-video retrieval (TVR) systems often suffer from visual-linguistic biases present in datasets, which cause pre-trained vision-language models to overlook key details. To address this, we propose BiMa, a novel framework designed to mitigate biases in both visual and textual representations. Our approach begins by generating scene elements that characterize each video by identifying relevant entities/objects and activities. For visual debiasing, we integrate these scene elements into the video embeddings, enhancing them to emphasize fine-grained and salient details. For textual debiasing, we"},"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":"2506.03589","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-06-04T05:40:54Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"7d7793c02c64abd42b51c2afa53f80c7cebcbd2702d018bc15191d98a90efc36","abstract_canon_sha256":"9ccbb55c1c3d0c2555d589ba691927fb7e0c07257c72f06ec7dbaa8e76485a7d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:32:29.997186Z","signature_b64":"DazENtPElUH5xrZayUjbKB8imEsbB5K+OekhDN9ceodCSy4pLBPbFSOMoosPWy3S5ARTyzd1FPTt+QGTEo3ABw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7075807aabd015bf6ddb69302786bf72bfaa2a62d36ce1c8b94fae551545b1e0","last_reissued_at":"2026-07-05T11:32:29.996710Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:32:29.996710Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"BiMa: Towards Biases Mitigation for Text-Video Retrieval via Scene Element Guidance","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.CV","authors_text":"Anh Nguyen, Huy Le, Ngan Le, Nhat Chung, Tung Kieu","submitted_at":"2025-06-04T05:40:54Z","abstract_excerpt":"Text-video retrieval (TVR) systems often suffer from visual-linguistic biases present in datasets, which cause pre-trained vision-language models to overlook key details. To address this, we propose BiMa, a novel framework designed to mitigate biases in both visual and textual representations. Our approach begins by generating scene elements that characterize each video by identifying relevant entities/objects and activities. For visual debiasing, we integrate these scene elements into the video embeddings, enhancing them to emphasize fine-grained and salient details. For textual debiasing, we"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.03589","kind":"arxiv","version":3},"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/2506.03589/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":"2506.03589","created_at":"2026-07-05T11:32:29.996767+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.03589v3","created_at":"2026-07-05T11:32:29.996767+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.03589","created_at":"2026-07-05T11:32:29.996767+00:00"},{"alias_kind":"pith_short_12","alias_value":"OB2YA6VL2AK3","created_at":"2026-07-05T11:32:29.996767+00:00"},{"alias_kind":"pith_short_16","alias_value":"OB2YA6VL2AK363O3","created_at":"2026-07-05T11:32:29.996767+00:00"},{"alias_kind":"pith_short_8","alias_value":"OB2YA6VL","created_at":"2026-07-05T11:32:29.996767+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/OB2YA6VL2AK363O3NEYCPBV7OK","json":"https://pith.science/pith/OB2YA6VL2AK363O3NEYCPBV7OK.json","graph_json":"https://pith.science/api/pith-number/OB2YA6VL2AK363O3NEYCPBV7OK/graph.json","events_json":"https://pith.science/api/pith-number/OB2YA6VL2AK363O3NEYCPBV7OK/events.json","paper":"https://pith.science/paper/OB2YA6VL"},"agent_actions":{"view_html":"https://pith.science/pith/OB2YA6VL2AK363O3NEYCPBV7OK","download_json":"https://pith.science/pith/OB2YA6VL2AK363O3NEYCPBV7OK.json","view_paper":"https://pith.science/paper/OB2YA6VL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.03589&json=true","fetch_graph":"https://pith.science/api/pith-number/OB2YA6VL2AK363O3NEYCPBV7OK/graph.json","fetch_events":"https://pith.science/api/pith-number/OB2YA6VL2AK363O3NEYCPBV7OK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OB2YA6VL2AK363O3NEYCPBV7OK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OB2YA6VL2AK363O3NEYCPBV7OK/action/storage_attestation","attest_author":"https://pith.science/pith/OB2YA6VL2AK363O3NEYCPBV7OK/action/author_attestation","sign_citation":"https://pith.science/pith/OB2YA6VL2AK363O3NEYCPBV7OK/action/citation_signature","submit_replication":"https://pith.science/pith/OB2YA6VL2AK363O3NEYCPBV7OK/action/replication_record"}},"created_at":"2026-07-05T11:32:29.996767+00:00","updated_at":"2026-07-05T11:32:29.996767+00:00"}