{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:XS2CLQEWCB3HWP2C75HBPKX4AA","short_pith_number":"pith:XS2CLQEW","schema_version":"1.0","canonical_sha256":"bcb425c09610767b3f42ff4e17aafc0019e8c0e46a6f9543e0eb38b84ee1390a","source":{"kind":"arxiv","id":"2402.19479","version":1},"attestation_state":"computed","paper":{"title":"Panda-70M: Captioning 70M Videos with Multiple Cross-Modality Teachers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Aliaksandr Siarohin, Byung Eun Jeon, Ekaterina Deyneka, Hsiang-wei Chao, Hsin-Ying Lee, Jian Ren, Ming-Hsuan Yang, Sergey Tulyakov, Tsai-Shien Chen, Willi Menapace, Yuwei Fang","submitted_at":"2024-02-29T18:59:50Z","abstract_excerpt":"The quality of the data and annotation upper-bounds the quality of a downstream model. While there exist large text corpora and image-text pairs, high-quality video-text data is much harder to collect. First of all, manual labeling is more time-consuming, as it requires an annotator to watch an entire video. Second, videos have a temporal dimension, consisting of several scenes stacked together, and showing multiple actions. Accordingly, to establish a video dataset with high-quality captions, we propose an automatic approach leveraging multimodal inputs, such as textual video description, sub"},"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":"2402.19479","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-02-29T18:59:50Z","cross_cats_sorted":[],"title_canon_sha256":"f28411df0063f91af0a7db80dcb2b5b56fe37ce53382da224f879c06f048e479","abstract_canon_sha256":"8d8d34da8d6b8d5fa0fe7b2d568543749c150288022d328e78de7ab0f27e177a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:50:45.798671Z","signature_b64":"1j7NSozct5xL5JQI9LaBmaOWcBn4ahXQI0UktvfAeNKZy/j/Hki5HYGE7nWQom+z0n5ie75PK33C+Fx9iwCUCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bcb425c09610767b3f42ff4e17aafc0019e8c0e46a6f9543e0eb38b84ee1390a","last_reissued_at":"2026-07-05T07:50:45.798166Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:50:45.798166Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Panda-70M: Captioning 70M Videos with Multiple Cross-Modality Teachers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Aliaksandr Siarohin, Byung Eun Jeon, Ekaterina Deyneka, Hsiang-wei Chao, Hsin-Ying Lee, Jian Ren, Ming-Hsuan Yang, Sergey Tulyakov, Tsai-Shien Chen, Willi Menapace, Yuwei Fang","submitted_at":"2024-02-29T18:59:50Z","abstract_excerpt":"The quality of the data and annotation upper-bounds the quality of a downstream model. While there exist large text corpora and image-text pairs, high-quality video-text data is much harder to collect. First of all, manual labeling is more time-consuming, as it requires an annotator to watch an entire video. Second, videos have a temporal dimension, consisting of several scenes stacked together, and showing multiple actions. Accordingly, to establish a video dataset with high-quality captions, we propose an automatic approach leveraging multimodal inputs, such as textual video description, sub"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.19479","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/2402.19479/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":"2402.19479","created_at":"2026-07-05T07:50:45.798219+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.19479v1","created_at":"2026-07-05T07:50:45.798219+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.19479","created_at":"2026-07-05T07:50:45.798219+00:00"},{"alias_kind":"pith_short_12","alias_value":"XS2CLQEWCB3H","created_at":"2026-07-05T07:50:45.798219+00:00"},{"alias_kind":"pith_short_16","alias_value":"XS2CLQEWCB3HWP2C","created_at":"2026-07-05T07:50:45.798219+00:00"},{"alias_kind":"pith_short_8","alias_value":"XS2CLQEW","created_at":"2026-07-05T07:50:45.798219+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2406.03520","citing_title":"VideoPhy: Evaluating Physical Commonsense for Video Generation","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2501.13106","citing_title":"VideoLLaMA 3: Frontier Multimodal Foundation Models for Image and Video Understanding","ref_index":113,"is_internal_anchor":false},{"citing_arxiv_id":"2410.02713","citing_title":"LLaVA-Video: Video Instruction Tuning With Synthetic Data","ref_index":178,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XS2CLQEWCB3HWP2C75HBPKX4AA","json":"https://pith.science/pith/XS2CLQEWCB3HWP2C75HBPKX4AA.json","graph_json":"https://pith.science/api/pith-number/XS2CLQEWCB3HWP2C75HBPKX4AA/graph.json","events_json":"https://pith.science/api/pith-number/XS2CLQEWCB3HWP2C75HBPKX4AA/events.json","paper":"https://pith.science/paper/XS2CLQEW"},"agent_actions":{"view_html":"https://pith.science/pith/XS2CLQEWCB3HWP2C75HBPKX4AA","download_json":"https://pith.science/pith/XS2CLQEWCB3HWP2C75HBPKX4AA.json","view_paper":"https://pith.science/paper/XS2CLQEW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.19479&json=true","fetch_graph":"https://pith.science/api/pith-number/XS2CLQEWCB3HWP2C75HBPKX4AA/graph.json","fetch_events":"https://pith.science/api/pith-number/XS2CLQEWCB3HWP2C75HBPKX4AA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XS2CLQEWCB3HWP2C75HBPKX4AA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XS2CLQEWCB3HWP2C75HBPKX4AA/action/storage_attestation","attest_author":"https://pith.science/pith/XS2CLQEWCB3HWP2C75HBPKX4AA/action/author_attestation","sign_citation":"https://pith.science/pith/XS2CLQEWCB3HWP2C75HBPKX4AA/action/citation_signature","submit_replication":"https://pith.science/pith/XS2CLQEWCB3HWP2C75HBPKX4AA/action/replication_record"}},"created_at":"2026-07-05T07:50:45.798219+00:00","updated_at":"2026-07-05T07:50:45.798219+00:00"}