{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:LTXSYXSGEN3V66M352E2ECHDS6","short_pith_number":"pith:LTXSYXSG","schema_version":"1.0","canonical_sha256":"5cef2c5e4623775f799bee89a208e397b63b372c05863d1973e8a9ce42cc3623","source":{"kind":"arxiv","id":"2410.19702","version":2},"attestation_state":"computed","paper":{"title":"TimeSuite: Improving MLLMs for Long Video Understanding via Grounded Tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.MM"],"primary_cat":"cs.CV","authors_text":"Chenting Wang, Kunchang Li, Limin Wang, Songze Li, Tianxiang Jiang, Xiangyu Zeng, Xinhao Li, Yali Wang, Yansong Shi, Yi Wang, Yu Qiao, Zhengrong Yue, Ziang Yan","submitted_at":"2024-10-25T17:19:55Z","abstract_excerpt":"Multimodal Large Language Models (MLLMs) have demonstrated impressive performance in short video understanding. However, understanding long-form videos still remains challenging for MLLMs. This paper proposes TimeSuite, a collection of new designs to adapt the existing short-form video MLLMs for long video understanding, including a simple yet efficient framework to process long video sequence, a high-quality video dataset for grounded tuning of MLLMs, and a carefully-designed instruction tuning task to explicitly incorporate the grounding supervision in the traditional QA format. Specifically"},"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":"2410.19702","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-10-25T17:19:55Z","cross_cats_sorted":["cs.AI","cs.MM"],"title_canon_sha256":"c2df90cb5d7e916075b61ec13d7fb53b88d7e9c24cf029d7023b9d5744f17a7a","abstract_canon_sha256":"fecfb611ff3a54d914707e1d6ee2317d69d22bb70b6d21f96fe945e399cd73e1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:13:04.935032Z","signature_b64":"uJtTzef6GF2D1ijTxf5nkfeGRJbhfbAZ/jMFI05kd9Ef0KwUpVcIekLWkFnHB32FnLyTi/J65XzUxPPN1nb8DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5cef2c5e4623775f799bee89a208e397b63b372c05863d1973e8a9ce42cc3623","last_reissued_at":"2026-07-05T10:13:04.934572Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:13:04.934572Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"TimeSuite: Improving MLLMs for Long Video Understanding via Grounded Tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.MM"],"primary_cat":"cs.CV","authors_text":"Chenting Wang, Kunchang Li, Limin Wang, Songze Li, Tianxiang Jiang, Xiangyu Zeng, Xinhao Li, Yali Wang, Yansong Shi, Yi Wang, Yu Qiao, Zhengrong Yue, Ziang Yan","submitted_at":"2024-10-25T17:19:55Z","abstract_excerpt":"Multimodal Large Language Models (MLLMs) have demonstrated impressive performance in short video understanding. However, understanding long-form videos still remains challenging for MLLMs. This paper proposes TimeSuite, a collection of new designs to adapt the existing short-form video MLLMs for long video understanding, including a simple yet efficient framework to process long video sequence, a high-quality video dataset for grounded tuning of MLLMs, and a carefully-designed instruction tuning task to explicitly incorporate the grounding supervision in the traditional QA format. Specifically"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.19702","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/2410.19702/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":"2410.19702","created_at":"2026-07-05T10:13:04.934634+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.19702v2","created_at":"2026-07-05T10:13:04.934634+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.19702","created_at":"2026-07-05T10:13:04.934634+00:00"},{"alias_kind":"pith_short_12","alias_value":"LTXSYXSGEN3V","created_at":"2026-07-05T10:13:04.934634+00:00"},{"alias_kind":"pith_short_16","alias_value":"LTXSYXSGEN3V66M3","created_at":"2026-07-05T10:13:04.934634+00:00"},{"alias_kind":"pith_short_8","alias_value":"LTXSYXSG","created_at":"2026-07-05T10:13:04.934634+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":13,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.06294","citing_title":"Towards One-to-Many Temporal Grounding","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2604.25886","citing_title":"MarkIt: Training-Free Visual Markers for Precise Video Temporal Grounding","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2605.14733","citing_title":"Video-Zero: Self-Evolution Video Understanding","ref_index":51,"is_internal_anchor":false},{"citing_arxiv_id":"2605.21954","citing_title":"MLLMs Know When Before Speaking: Revealing and Recovering Temporal Grounding via Attention Cues","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2501.00574","citing_title":"VideoChat-Flash: Hierarchical Compression for Long-Context Video Modeling","ref_index":64,"is_internal_anchor":false},{"citing_arxiv_id":"2512.03963","citing_title":"TempR1: Improving Temporal Understanding of MLLMs via Temporal-Aware Multi-Task Reinforcement Learning","ref_index":67,"is_internal_anchor":false},{"citing_arxiv_id":"2512.03043","citing_title":"OneThinker: All-in-one Reasoning Model for Image and Video","ref_index":72,"is_internal_anchor":false},{"citing_arxiv_id":"2602.02994","citing_title":"Video-OPD: Efficient Post-Training of Multimodal Large Language Models for Temporal Video Grounding via On-Policy Distillation","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2504.06958","citing_title":"VideoChat-R1: Enhancing Spatio-Temporal Perception via Reinforcement Fine-Tuning","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13803","citing_title":"EvoGround: Self-Evolving Video Agents for Video Temporal Grounding","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2604.25886","citing_title":"MarkIt: Training-Free Visual Markers for Precise Video Temporal Grounding","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2604.09037","citing_title":"SiMing-Bench: Evaluating Procedural Correctness from Continuous Interactions in Clinical Skill Videos","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2604.08014","citing_title":"Bridging Time and Space: Decoupled Spatio-Temporal Alignment for Video Grounding","ref_index":71,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LTXSYXSGEN3V66M352E2ECHDS6","json":"https://pith.science/pith/LTXSYXSGEN3V66M352E2ECHDS6.json","graph_json":"https://pith.science/api/pith-number/LTXSYXSGEN3V66M352E2ECHDS6/graph.json","events_json":"https://pith.science/api/pith-number/LTXSYXSGEN3V66M352E2ECHDS6/events.json","paper":"https://pith.science/paper/LTXSYXSG"},"agent_actions":{"view_html":"https://pith.science/pith/LTXSYXSGEN3V66M352E2ECHDS6","download_json":"https://pith.science/pith/LTXSYXSGEN3V66M352E2ECHDS6.json","view_paper":"https://pith.science/paper/LTXSYXSG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.19702&json=true","fetch_graph":"https://pith.science/api/pith-number/LTXSYXSGEN3V66M352E2ECHDS6/graph.json","fetch_events":"https://pith.science/api/pith-number/LTXSYXSGEN3V66M352E2ECHDS6/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LTXSYXSGEN3V66M352E2ECHDS6/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LTXSYXSGEN3V66M352E2ECHDS6/action/storage_attestation","attest_author":"https://pith.science/pith/LTXSYXSGEN3V66M352E2ECHDS6/action/author_attestation","sign_citation":"https://pith.science/pith/LTXSYXSGEN3V66M352E2ECHDS6/action/citation_signature","submit_replication":"https://pith.science/pith/LTXSYXSGEN3V66M352E2ECHDS6/action/replication_record"}},"created_at":"2026-07-05T10:13:04.934634+00:00","updated_at":"2026-07-05T10:13:04.934634+00:00"}