{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:LGE7PQ342OR7AW7XYY7PZUA7DV","short_pith_number":"pith:LGE7PQ34","schema_version":"1.0","canonical_sha256":"5989f7c37cd3a3f05bf7c63efcd01f1d4a2ab3ba1da8cfb53f906749b922da33","source":{"kind":"arxiv","id":"2301.01953","version":3},"attestation_state":"computed","paper":{"title":"Learning Trajectory-Word Alignments for Video-Language Tasks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chenliang Li, Fei Huang, Haiyang Xu, Hanwang Zhang, Ming Yan, Qinghao Ye, Songfang Huang, Xu Yang, Yu Zhang, Zhangzikang Li","submitted_at":"2023-01-05T08:21:01Z","abstract_excerpt":"In a video, an object usually appears as the trajectory, i.e., it spans over a few spatial but longer temporal patches, that contains abundant spatiotemporal contexts. However, modern Video-Language BERTs (VDL-BERTs) neglect this trajectory characteristic that they usually follow image-language BERTs (IL-BERTs) to deploy the patch-to-word (P2W) attention that may over-exploit trivial spatial contexts and neglect significant temporal contexts. To amend this, we propose a novel TW-BERT to learn Trajectory-Word alignment by a newly designed trajectory-to-word (T2W) attention for solving video-lan"},"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":"2301.01953","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2023-01-05T08:21:01Z","cross_cats_sorted":[],"title_canon_sha256":"5559cdc2e7bf9438bb10efb47f2c07d7018a7aaababed8508638d1671bb576b6","abstract_canon_sha256":"43adb24f2f1c4c80e11ff73ee54cfbcfd4b191735f40fe3b25f6f4f3b6895e58"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:49:32.253478Z","signature_b64":"IKjeUlVxD2qHOdMy624x2Ct3TZ+TI2fEsXWklh0Ep9XU8LTJss58qkfil6uQrS2Axlopd9o/f4NhKDesGdYcAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5989f7c37cd3a3f05bf7c63efcd01f1d4a2ab3ba1da8cfb53f906749b922da33","last_reissued_at":"2026-07-05T05:49:32.253101Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:49:32.253101Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Learning Trajectory-Word Alignments for Video-Language Tasks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chenliang Li, Fei Huang, Haiyang Xu, Hanwang Zhang, Ming Yan, Qinghao Ye, Songfang Huang, Xu Yang, Yu Zhang, Zhangzikang Li","submitted_at":"2023-01-05T08:21:01Z","abstract_excerpt":"In a video, an object usually appears as the trajectory, i.e., it spans over a few spatial but longer temporal patches, that contains abundant spatiotemporal contexts. However, modern Video-Language BERTs (VDL-BERTs) neglect this trajectory characteristic that they usually follow image-language BERTs (IL-BERTs) to deploy the patch-to-word (P2W) attention that may over-exploit trivial spatial contexts and neglect significant temporal contexts. To amend this, we propose a novel TW-BERT to learn Trajectory-Word alignment by a newly designed trajectory-to-word (T2W) attention for solving video-lan"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2301.01953","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/2301.01953/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":"2301.01953","created_at":"2026-07-05T05:49:32.253153+00:00"},{"alias_kind":"arxiv_version","alias_value":"2301.01953v3","created_at":"2026-07-05T05:49:32.253153+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2301.01953","created_at":"2026-07-05T05:49:32.253153+00:00"},{"alias_kind":"pith_short_12","alias_value":"LGE7PQ342OR7","created_at":"2026-07-05T05:49:32.253153+00:00"},{"alias_kind":"pith_short_16","alias_value":"LGE7PQ342OR7AW7X","created_at":"2026-07-05T05:49:32.253153+00:00"},{"alias_kind":"pith_short_8","alias_value":"LGE7PQ34","created_at":"2026-07-05T05:49:32.253153+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/LGE7PQ342OR7AW7XYY7PZUA7DV","json":"https://pith.science/pith/LGE7PQ342OR7AW7XYY7PZUA7DV.json","graph_json":"https://pith.science/api/pith-number/LGE7PQ342OR7AW7XYY7PZUA7DV/graph.json","events_json":"https://pith.science/api/pith-number/LGE7PQ342OR7AW7XYY7PZUA7DV/events.json","paper":"https://pith.science/paper/LGE7PQ34"},"agent_actions":{"view_html":"https://pith.science/pith/LGE7PQ342OR7AW7XYY7PZUA7DV","download_json":"https://pith.science/pith/LGE7PQ342OR7AW7XYY7PZUA7DV.json","view_paper":"https://pith.science/paper/LGE7PQ34","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2301.01953&json=true","fetch_graph":"https://pith.science/api/pith-number/LGE7PQ342OR7AW7XYY7PZUA7DV/graph.json","fetch_events":"https://pith.science/api/pith-number/LGE7PQ342OR7AW7XYY7PZUA7DV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LGE7PQ342OR7AW7XYY7PZUA7DV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LGE7PQ342OR7AW7XYY7PZUA7DV/action/storage_attestation","attest_author":"https://pith.science/pith/LGE7PQ342OR7AW7XYY7PZUA7DV/action/author_attestation","sign_citation":"https://pith.science/pith/LGE7PQ342OR7AW7XYY7PZUA7DV/action/citation_signature","submit_replication":"https://pith.science/pith/LGE7PQ342OR7AW7XYY7PZUA7DV/action/replication_record"}},"created_at":"2026-07-05T05:49:32.253153+00:00","updated_at":"2026-07-05T05:49:32.253153+00:00"}