{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:Y2JXR346O7RXHNFYN6QL3GAKRX","short_pith_number":"pith:Y2JXR346","schema_version":"1.0","canonical_sha256":"c69378ef9e77e373b4b86fa0bd980a8dfa270bc4bd5dc09f662e4483bb5dbdd5","source":{"kind":"arxiv","id":"2410.11417","version":1},"attestation_state":"computed","paper":{"title":"VidCompress: Memory-Enhanced Temporal Compression for Video Understanding in Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.MM"],"primary_cat":"cs.CV","authors_text":"Lin Ma, Xiaohan Lan, Yitian Yuan, Zequn Jie","submitted_at":"2024-10-15T09:07:25Z","abstract_excerpt":"Video-based multimodal large language models (Video-LLMs) possess significant potential for video understanding tasks. However, most Video-LLMs treat videos as a sequential set of individual frames, which results in insufficient temporal-spatial interaction that hinders fine-grained comprehension and difficulty in processing longer videos due to limited visual token capacity. To address these challenges, we propose VidCompress, a novel Video-LLM featuring memory-enhanced temporal compression. VidCompress employs a dual-compressor approach: a memory-enhanced compressor captures both short-term "},"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":"2410.11417","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-10-15T09:07:25Z","cross_cats_sorted":["cs.MM"],"title_canon_sha256":"54beeaa452a1f2788a260c97d9b96841fb7106f884010f994c40745941d90603","abstract_canon_sha256":"d3c3ee92076b3ece90b4c86972e6a8f36621ebcc50405416f3915804ed3d49d8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:20:53.224777Z","signature_b64":"h+JFVYVL5U6EFh/gCI2QAXoNeWKZs1wF6rga4UGgJrHcsB50mktZFQCsk1wPNtYebE3B6xUY+nwYFiW7GwdKBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c69378ef9e77e373b4b86fa0bd980a8dfa270bc4bd5dc09f662e4483bb5dbdd5","last_reissued_at":"2026-07-05T09:20:53.224289Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:20:53.224289Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"VidCompress: Memory-Enhanced Temporal Compression for Video Understanding in Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.MM"],"primary_cat":"cs.CV","authors_text":"Lin Ma, Xiaohan Lan, Yitian Yuan, Zequn Jie","submitted_at":"2024-10-15T09:07:25Z","abstract_excerpt":"Video-based multimodal large language models (Video-LLMs) possess significant potential for video understanding tasks. However, most Video-LLMs treat videos as a sequential set of individual frames, which results in insufficient temporal-spatial interaction that hinders fine-grained comprehension and difficulty in processing longer videos due to limited visual token capacity. To address these challenges, we propose VidCompress, a novel Video-LLM featuring memory-enhanced temporal compression. VidCompress employs a dual-compressor approach: a memory-enhanced compressor captures both short-term "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.11417","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/2410.11417/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":"2410.11417","created_at":"2026-07-05T09:20:53.224348+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.11417v1","created_at":"2026-07-05T09:20:53.224348+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.11417","created_at":"2026-07-05T09:20:53.224348+00:00"},{"alias_kind":"pith_short_12","alias_value":"Y2JXR346O7RX","created_at":"2026-07-05T09:20:53.224348+00:00"},{"alias_kind":"pith_short_16","alias_value":"Y2JXR346O7RXHNFY","created_at":"2026-07-05T09:20:53.224348+00:00"},{"alias_kind":"pith_short_8","alias_value":"Y2JXR346","created_at":"2026-07-05T09:20:53.224348+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.07433","citing_title":"Watch, Remember, Reason: Human-View Video Understanding with MLLMs","ref_index":149,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/Y2JXR346O7RXHNFYN6QL3GAKRX","json":"https://pith.science/pith/Y2JXR346O7RXHNFYN6QL3GAKRX.json","graph_json":"https://pith.science/api/pith-number/Y2JXR346O7RXHNFYN6QL3GAKRX/graph.json","events_json":"https://pith.science/api/pith-number/Y2JXR346O7RXHNFYN6QL3GAKRX/events.json","paper":"https://pith.science/paper/Y2JXR346"},"agent_actions":{"view_html":"https://pith.science/pith/Y2JXR346O7RXHNFYN6QL3GAKRX","download_json":"https://pith.science/pith/Y2JXR346O7RXHNFYN6QL3GAKRX.json","view_paper":"https://pith.science/paper/Y2JXR346","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.11417&json=true","fetch_graph":"https://pith.science/api/pith-number/Y2JXR346O7RXHNFYN6QL3GAKRX/graph.json","fetch_events":"https://pith.science/api/pith-number/Y2JXR346O7RXHNFYN6QL3GAKRX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/Y2JXR346O7RXHNFYN6QL3GAKRX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/Y2JXR346O7RXHNFYN6QL3GAKRX/action/storage_attestation","attest_author":"https://pith.science/pith/Y2JXR346O7RXHNFYN6QL3GAKRX/action/author_attestation","sign_citation":"https://pith.science/pith/Y2JXR346O7RXHNFYN6QL3GAKRX/action/citation_signature","submit_replication":"https://pith.science/pith/Y2JXR346O7RXHNFYN6QL3GAKRX/action/replication_record"}},"created_at":"2026-07-05T09:20:53.224348+00:00","updated_at":"2026-07-05T09:20:53.224348+00:00"}