{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:OJYTNIZ7AQ6XLLFDPXNFIS7UDA","short_pith_number":"pith:OJYTNIZ7","schema_version":"1.0","canonical_sha256":"727136a33f043d75aca37dda544bf41831b9b1ede750d236044751b29b224dc9","source":{"kind":"arxiv","id":"2407.00603","version":2},"attestation_state":"computed","paper":{"title":"Hierarchical Memory for Long Video QA","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Haoji Zhang, Jiashi Feng, Jifeng Dai, Xiaojie Jin, Yansong Tang, Yiqin Wang, Yong Liu","submitted_at":"2024-06-30T06:08:12Z","abstract_excerpt":"This paper describes our champion solution to the LOVEU Challenge @ CVPR'24, Track 1 (Long Video VQA). Processing long sequences of visual tokens is computationally expensive and memory-intensive, making long video question-answering a challenging task. The key is to compress visual tokens effectively, reducing memory footprint and decoding latency, while preserving the essential information for accurate question-answering. We adopt a hierarchical memory mechanism named STAR Memory, proposed in Flash-VStream, that is capable of processing long videos with limited GPU memory (VRAM). We further "},"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":"2407.00603","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2024-06-30T06:08:12Z","cross_cats_sorted":[],"title_canon_sha256":"f0c5b40d9efbef27615cda6cd7b5c46a047825f1ce39132d1e0bb40650e5d3cb","abstract_canon_sha256":"f8efbe28c031fd5b72760113f5f840c56b006e2e4e2debedde2ca7a076c353e5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:49:07.285886Z","signature_b64":"J7LkDYhE5oNaSjrytfyFw/dO1RZzRYAGTgjLG01ehRe39jRt1KggrQvPc4kJVPtBk5gxhviw0VY0Q90taCF/AA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"727136a33f043d75aca37dda544bf41831b9b1ede750d236044751b29b224dc9","last_reissued_at":"2026-07-05T09:49:07.285336Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:49:07.285336Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Hierarchical Memory for Long Video QA","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Haoji Zhang, Jiashi Feng, Jifeng Dai, Xiaojie Jin, Yansong Tang, Yiqin Wang, Yong Liu","submitted_at":"2024-06-30T06:08:12Z","abstract_excerpt":"This paper describes our champion solution to the LOVEU Challenge @ CVPR'24, Track 1 (Long Video VQA). Processing long sequences of visual tokens is computationally expensive and memory-intensive, making long video question-answering a challenging task. The key is to compress visual tokens effectively, reducing memory footprint and decoding latency, while preserving the essential information for accurate question-answering. We adopt a hierarchical memory mechanism named STAR Memory, proposed in Flash-VStream, that is capable of processing long videos with limited GPU memory (VRAM). We further "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.00603","kind":"arxiv","version":2},"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/2407.00603/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":"2407.00603","created_at":"2026-07-05T09:49:07.285410+00:00"},{"alias_kind":"arxiv_version","alias_value":"2407.00603v2","created_at":"2026-07-05T09:49:07.285410+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.00603","created_at":"2026-07-05T09:49:07.285410+00:00"},{"alias_kind":"pith_short_12","alias_value":"OJYTNIZ7AQ6X","created_at":"2026-07-05T09:49:07.285410+00:00"},{"alias_kind":"pith_short_16","alias_value":"OJYTNIZ7AQ6XLLFD","created_at":"2026-07-05T09:49:07.285410+00:00"},{"alias_kind":"pith_short_8","alias_value":"OJYTNIZ7","created_at":"2026-07-05T09:49:07.285410+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.05736","citing_title":"VTI-CoT: Visual-Textual Interleaved Chain of Thought for Video Reasoning","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17921","citing_title":"An Efficient Streaming Video Understanding Framework with Agentic Control","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2505.18719","citing_title":"VLA-RL: Towards Masterful and General Robotic Manipulation with Scalable Reinforcement Learning","ref_index":73,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/OJYTNIZ7AQ6XLLFDPXNFIS7UDA","json":"https://pith.science/pith/OJYTNIZ7AQ6XLLFDPXNFIS7UDA.json","graph_json":"https://pith.science/api/pith-number/OJYTNIZ7AQ6XLLFDPXNFIS7UDA/graph.json","events_json":"https://pith.science/api/pith-number/OJYTNIZ7AQ6XLLFDPXNFIS7UDA/events.json","paper":"https://pith.science/paper/OJYTNIZ7"},"agent_actions":{"view_html":"https://pith.science/pith/OJYTNIZ7AQ6XLLFDPXNFIS7UDA","download_json":"https://pith.science/pith/OJYTNIZ7AQ6XLLFDPXNFIS7UDA.json","view_paper":"https://pith.science/paper/OJYTNIZ7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2407.00603&json=true","fetch_graph":"https://pith.science/api/pith-number/OJYTNIZ7AQ6XLLFDPXNFIS7UDA/graph.json","fetch_events":"https://pith.science/api/pith-number/OJYTNIZ7AQ6XLLFDPXNFIS7UDA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OJYTNIZ7AQ6XLLFDPXNFIS7UDA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OJYTNIZ7AQ6XLLFDPXNFIS7UDA/action/storage_attestation","attest_author":"https://pith.science/pith/OJYTNIZ7AQ6XLLFDPXNFIS7UDA/action/author_attestation","sign_citation":"https://pith.science/pith/OJYTNIZ7AQ6XLLFDPXNFIS7UDA/action/citation_signature","submit_replication":"https://pith.science/pith/OJYTNIZ7AQ6XLLFDPXNFIS7UDA/action/replication_record"}},"created_at":"2026-07-05T09:49:07.285410+00:00","updated_at":"2026-07-05T09:49:07.285410+00:00"}