{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:SCPKYW2DBYLL4BMWFHQ6PZDXAG","short_pith_number":"pith:SCPKYW2D","schema_version":"1.0","canonical_sha256":"909eac5b430e16be059629e1e7e4770196c296325a4c4aa46062cfeb8376a978","source":{"kind":"arxiv","id":"2505.04512","version":2},"attestation_state":"computed","paper":{"title":"HunyuanCustom: A Multimodal-Driven Architecture for Customized Video Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Qinglin Lu, Qin Lin, Sen Liang, Teng Hu, Yuan Zhou, Zhengguang Zhou, Zhentao Yu","submitted_at":"2025-05-07T15:33:18Z","abstract_excerpt":"Customized video generation aims to produce videos featuring specific subjects under flexible user-defined conditions, yet existing methods often struggle with identity consistency and limited input modalities. In this paper, we propose HunyuanCustom, a multi-modal customized video generation framework that emphasizes subject consistency while supporting image, audio, video, and text conditions. Built upon HunyuanVideo, our model first addresses the image-text conditioned generation task by introducing a text-image fusion module based on LLaVA for enhanced multi-modal understanding, along with"},"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":"2505.04512","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-05-07T15:33:18Z","cross_cats_sorted":[],"title_canon_sha256":"b1a3f7a332ce3878a5628023284a380795bbce321bd3af781fad78d8eebc43cf","abstract_canon_sha256":"52057e406c51d9df0eeac35b945b4acb57037c0bac81498a89e1150597f722cb"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:00:10.282791Z","signature_b64":"gR/vZI/7AqoZ5Qh51Gdqi00sBlGDeX546HNdtYCEo/Fmg5DWClxOFMcolsHdI8Dyb8fE8NRX7Ahij3nSEpCSDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"909eac5b430e16be059629e1e7e4770196c296325a4c4aa46062cfeb8376a978","last_reissued_at":"2026-07-05T11:00:10.282142Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:00:10.282142Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"HunyuanCustom: A Multimodal-Driven Architecture for Customized Video Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Qinglin Lu, Qin Lin, Sen Liang, Teng Hu, Yuan Zhou, Zhengguang Zhou, Zhentao Yu","submitted_at":"2025-05-07T15:33:18Z","abstract_excerpt":"Customized video generation aims to produce videos featuring specific subjects under flexible user-defined conditions, yet existing methods often struggle with identity consistency and limited input modalities. In this paper, we propose HunyuanCustom, a multi-modal customized video generation framework that emphasizes subject consistency while supporting image, audio, video, and text conditions. Built upon HunyuanVideo, our model first addresses the image-text conditioned generation task by introducing a text-image fusion module based on LLaVA for enhanced multi-modal understanding, along with"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.04512","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/2505.04512/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":"2505.04512","created_at":"2026-07-05T11:00:10.282221+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.04512v2","created_at":"2026-07-05T11:00:10.282221+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.04512","created_at":"2026-07-05T11:00:10.282221+00:00"},{"alias_kind":"pith_short_12","alias_value":"SCPKYW2DBYLL","created_at":"2026-07-05T11:00:10.282221+00:00"},{"alias_kind":"pith_short_16","alias_value":"SCPKYW2DBYLL4BMW","created_at":"2026-07-05T11:00:10.282221+00:00"},{"alias_kind":"pith_short_8","alias_value":"SCPKYW2D","created_at":"2026-07-05T11:00:10.282221+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":18,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.26058","citing_title":"DomainShuttle: Freeform Open Domain Subject-driven Text-to-video Generation","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2606.11670","citing_title":"ARGUS: Stacked Multi-View Identity Mosaic Injection for Subject-Preserving Video Generation","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2606.10839","citing_title":"HarmoView: Harmonizing Multi-View Constraints for Identity-Consistent Video Generation","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2606.09639","citing_title":"CineDance: Towards Next-Generation Multi-Shot Long-Form Cinematic Audio-Video Generation","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2607.01222","citing_title":"Ink3D: Sculpting 3D Assets with Extremely Complex Textures via Video Generative Models","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02753","citing_title":"MetaWorld: Scaling Multi-Agent Video World Model from Single-view Video Data","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02441","citing_title":"Spatial-Temporal Decoupled Reference Conditioning for Identity-Preserving Text-to-Video Generation","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17488","citing_title":"Omni-Customizer: End-to-End MultiModal Customization for Joint Audio-Video Generation","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12088","citing_title":"UniCustom: Unified Visual Conditioning for Multi-Reference Image Generation","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12088","citing_title":"UniCustom: Unified Visual Conditioning for Multi-Reference Image Generation","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2605.04702","citing_title":"FaithfulFaces: Pose-Faithful Facial Identity Preservation for Text-to-Video Generation","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2604.19679","citing_title":"MMControl: Unified Multi-Modal Control for Joint Audio-Video Generation","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2604.11244","citing_title":"Script-a-Video: Deep Structured Audio-visual Captions via Factorized Streams and Relational Grounding","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2604.11804","citing_title":"OmniShow: Unifying Multimodal Conditions for Human-Object Interaction Video Generation","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2604.10030","citing_title":"Prompt Relay: Inference-Time Temporal Control for Multi-Event Video Generation","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06339","citing_title":"Evolution of Video Generative Foundations","ref_index":204,"is_internal_anchor":false},{"citing_arxiv_id":"2604.14556","citing_title":"Controllable Video Object Insertion via Multiview Priors","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2604.19473","citing_title":"TS-Attn: Temporal-wise Separable Attention for Multi-Event Video Generation","ref_index":10,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SCPKYW2DBYLL4BMWFHQ6PZDXAG","json":"https://pith.science/pith/SCPKYW2DBYLL4BMWFHQ6PZDXAG.json","graph_json":"https://pith.science/api/pith-number/SCPKYW2DBYLL4BMWFHQ6PZDXAG/graph.json","events_json":"https://pith.science/api/pith-number/SCPKYW2DBYLL4BMWFHQ6PZDXAG/events.json","paper":"https://pith.science/paper/SCPKYW2D"},"agent_actions":{"view_html":"https://pith.science/pith/SCPKYW2DBYLL4BMWFHQ6PZDXAG","download_json":"https://pith.science/pith/SCPKYW2DBYLL4BMWFHQ6PZDXAG.json","view_paper":"https://pith.science/paper/SCPKYW2D","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.04512&json=true","fetch_graph":"https://pith.science/api/pith-number/SCPKYW2DBYLL4BMWFHQ6PZDXAG/graph.json","fetch_events":"https://pith.science/api/pith-number/SCPKYW2DBYLL4BMWFHQ6PZDXAG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SCPKYW2DBYLL4BMWFHQ6PZDXAG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SCPKYW2DBYLL4BMWFHQ6PZDXAG/action/storage_attestation","attest_author":"https://pith.science/pith/SCPKYW2DBYLL4BMWFHQ6PZDXAG/action/author_attestation","sign_citation":"https://pith.science/pith/SCPKYW2DBYLL4BMWFHQ6PZDXAG/action/citation_signature","submit_replication":"https://pith.science/pith/SCPKYW2DBYLL4BMWFHQ6PZDXAG/action/replication_record"}},"created_at":"2026-07-05T11:00:10.282221+00:00","updated_at":"2026-07-05T11:00:10.282221+00:00"}