{"paper":{"title":"AuDirector: A Self-Reflective Closed-Loop Framework for Immersive Audio Storytelling","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"AuDirector's self-reflective closed-loop multi-agent framework produces long-form audio stories with greater structural coherence, emotional expressiveness, and acoustic fidelity than existing approaches.","cross_cats":[],"primary_cat":"cs.SD","authors_text":"Baoxiang Li, Chao Zhang, Wen Wu, Xuenan Xu, Yiming Ren, Ziyang Zhang","submitted_at":"2026-05-12T09:46:36Z","abstract_excerpt":"Despite advances in text and visual generation, creating coherent long-form audio narratives remains challenging. Existing frameworks often exhibit limitations such as mismatched character settings with voice performance, insufficient self-correction mechanisms, and limited human interactivity. To address these challenges, we propose AuDirector, a self-reflective closed-loop multi-agent framework. Specifically, it involves an Identity-Aware Pre-production mechanism that transforms narrative texts into character profiles and utterance-level emotional instructions to retrieve suitable voice cand"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Experiments demonstrate that AuDirector achieves superior performance compared to state-of-the-art baselines in structural coherence, emotional expressiveness, and acoustic fidelity.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the Identity-Aware Pre-production, Collaborative Synthesis and Correction, and Human-Guided Interactive Refinement modules can be integrated into a stable closed-loop system that consistently improves output quality without introducing new inconsistencies or requiring extensive manual tuning.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"AuDirector is a self-reflective closed-loop multi-agent framework that generates immersive audio narratives with improved structural coherence, emotional expressiveness, and acoustic fidelity via identity-aware voice adaptation and iterative correction.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"AuDirector's self-reflective closed-loop multi-agent framework produces long-form audio stories with greater structural coherence, emotional expressiveness, and acoustic fidelity than existing approaches.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"e381da6f9de5d6f5457413ddba178fea5fb691226c924601d33c88bcccb66262"},"source":{"id":"2605.11866","kind":"arxiv","version":2},"verdict":{"id":"78318dd6-62cf-40b8-8f33-bca880a48bea","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-13T05:01:02.479990Z","strongest_claim":"Experiments demonstrate that AuDirector achieves superior performance compared to state-of-the-art baselines in structural coherence, emotional expressiveness, and acoustic fidelity.","one_line_summary":"AuDirector is a self-reflective closed-loop multi-agent framework that generates immersive audio narratives with improved structural coherence, emotional expressiveness, and acoustic fidelity via identity-aware voice adaptation and iterative correction.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the Identity-Aware Pre-production, Collaborative Synthesis and Correction, and Human-Guided Interactive Refinement modules can be integrated into a stable closed-loop system that consistently improves output quality without introducing new inconsistencies or requiring extensive manual tuning.","pith_extraction_headline":"AuDirector's self-reflective closed-loop multi-agent framework produces long-form audio stories with greater structural coherence, emotional expressiveness, and acoustic fidelity than existing approaches."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.11866/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_title_agreement","ran_at":"2026-05-20T18:31:27.274027Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-20T12:46:25.031635Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"claim_evidence","ran_at":"2026-05-20T03:22:00.394214Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T11:36:33.471188Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"9dc6fd9bbb5f701929d8267b873efd802d17d71c0c6fcfa71fa50f5c6cb11218"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":2,"snapshot_sha256":"8d7dc417e9eeb3ce6e1e2c58164073a4f024006d90e6cc21497ed81c1ac6bf96"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}