{"paper":{"title":"Unison: Harmonizing Motion, Speech, and Sound for Human-Centric Audio-Video Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Unison is a unified framework that harmonizes motion, speech, and sound in human-centric video generation through explicit multimodal strategies.","cross_cats":["cs.GR","cs.MM","cs.SD"],"primary_cat":"cs.CV","authors_text":"Chi Zhang, Jiaxu Zhang, Quanyue Song, Shansong Liu, Shihao Cheng, Xiaolei Zhang, Xuelong Li, Zhigang Tu, Zhizhi Guo","submitted_at":"2026-05-09T06:32:54Z","abstract_excerpt":"Motion, speech, and sound effects are fundamental elements of human-centric videos, yet their heterogeneous temporal characteristics make joint generation highly challenging. Existing audio-video generation models often fail to maintain consistent alignment across these modalities, leading to noticeable mismatches between motion, speech, and environmental sounds. We present Unison, a unified framework that explicitly promotes coherence across the motion, speech, and sound modalities. Within the audio stream, Unison employs a semantic-guided harmonization strategy that decouples the generation "},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Unison achieves state-of-the-art performance in both audio perceptual quality and cross-modal synchronization, highlighting the importance of explicit multimodal harmonization in human-centric video generation.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the proposed semantic-guided harmonization strategy and bidirectional cross-modal forcing with decoupled denoising schedules will reliably improve coherence without introducing new artifacts or requiring extensive post-hoc tuning on specific datasets.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Unison introduces a unified framework using semantic-guided harmonization and bidirectional cross-modal forcing to generate human-centric videos with improved synchronization between motion, speech, and sound effects.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Unison is a unified framework that harmonizes motion, speech, and sound in human-centric video generation through explicit multimodal strategies.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"54dd8e8557b3b51e7d2c12ac7794564e5e260c77012ee7f1c072f4d935b545dc"},"source":{"id":"2605.08729","kind":"arxiv","version":2},"verdict":{"id":"df696f81-1a0b-48aa-a3b7-da2f9f70853d","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-12T02:23:25.744550Z","strongest_claim":"Unison achieves state-of-the-art performance in both audio perceptual quality and cross-modal synchronization, highlighting the importance of explicit multimodal harmonization in human-centric video generation.","one_line_summary":"Unison introduces a unified framework using semantic-guided harmonization and bidirectional cross-modal forcing to generate human-centric videos with improved synchronization between motion, speech, and sound effects.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the proposed semantic-guided harmonization strategy and bidirectional cross-modal forcing with decoupled denoising schedules will reliably improve coherence without introducing new artifacts or requiring extensive post-hoc tuning on specific datasets.","pith_extraction_headline":"Unison is a unified framework that harmonizes motion, speech, and sound in human-centric video generation through explicit multimodal strategies."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.08729/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"claim_evidence","ran_at":"2026-05-20T09:02:01.946657Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T22:35:38.357335Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_title_agreement","ran_at":"2026-05-19T14:31:17.607190Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T10:50:51.413104Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"0898b83cb733f80c4cd8792bde22780a797caf2f911c29f126496c569a0ed959"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":2,"snapshot_sha256":"8235d5616349662ad9114548ef4262d99d05bd2c63ae20dcdc17b9c234d0733f"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}