{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:ZI73ZZXFEI2HLXCVFMITQMZM2T","short_pith_number":"pith:ZI73ZZXF","schema_version":"1.0","canonical_sha256":"ca3fbce6e5223475dc552b1138332cd4d701ac2878d57bf0c27d5cf9b49c9864","source":{"kind":"arxiv","id":"2506.21619","version":2},"attestation_state":"computed","paper":{"title":"IndexTTS2: A Breakthrough in Emotionally Expressive and Duration-Controlled Auto-Regressive Zero-Shot Text-to-Speech","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.SD","eess.AS"],"primary_cat":"cs.CL","authors_text":"Jinchao Wang, Jingchen Shu, Siyi Zhou, Wei Deng, Xun Zhou, Yi He, Yiquan Zhou","submitted_at":"2025-06-23T08:33:40Z","abstract_excerpt":"Existing autoregressive large-scale text-to-speech (TTS) models have advantages in speech naturalness, but their token-by-token generation mechanism makes it difficult to precisely control the duration of synthesized speech. This becomes a significant limitation in applications requiring strict audio-visual synchronization, such as video dubbing. This paper introduces IndexTTS2, which proposes a novel, general, and autoregressive model-friendly method for speech duration control. The method supports two generation modes: one explicitly specifies the number of generated tokens to precisely cont"},"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":"2506.21619","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-06-23T08:33:40Z","cross_cats_sorted":["cs.AI","cs.SD","eess.AS"],"title_canon_sha256":"29cbb0e0a92f8a186bd6d22eafc280d7123bada383a0770ba5c52d614144e9ec","abstract_canon_sha256":"5b227c2324be2a3a70bcafdf8bb313f999d37b2cd4b502d2b8bae2ce1bc40464"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:03:49.476001Z","signature_b64":"3ozRWoEwEyBUpUAgynhkaxOlBVcyG3n3qjU4bW070Dgf+hswLS0uTVM6G+Df2iGbmY9pD/X+XUp4WPFYZmd1BQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ca3fbce6e5223475dc552b1138332cd4d701ac2878d57bf0c27d5cf9b49c9864","last_reissued_at":"2026-07-05T12:03:49.475486Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:03:49.475486Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"IndexTTS2: A Breakthrough in Emotionally Expressive and Duration-Controlled Auto-Regressive Zero-Shot Text-to-Speech","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.SD","eess.AS"],"primary_cat":"cs.CL","authors_text":"Jinchao Wang, Jingchen Shu, Siyi Zhou, Wei Deng, Xun Zhou, Yi He, Yiquan Zhou","submitted_at":"2025-06-23T08:33:40Z","abstract_excerpt":"Existing autoregressive large-scale text-to-speech (TTS) models have advantages in speech naturalness, but their token-by-token generation mechanism makes it difficult to precisely control the duration of synthesized speech. This becomes a significant limitation in applications requiring strict audio-visual synchronization, such as video dubbing. This paper introduces IndexTTS2, which proposes a novel, general, and autoregressive model-friendly method for speech duration control. The method supports two generation modes: one explicitly specifies the number of generated tokens to precisely cont"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.21619","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/2506.21619/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":"2506.21619","created_at":"2026-07-05T12:03:49.475546+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.21619v2","created_at":"2026-07-05T12:03:49.475546+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.21619","created_at":"2026-07-05T12:03:49.475546+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZI73ZZXFEI2H","created_at":"2026-07-05T12:03:49.475546+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZI73ZZXFEI2HLXCV","created_at":"2026-07-05T12:03:49.475546+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZI73ZZXF","created_at":"2026-07-05T12:03:49.475546+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":31,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.23190","citing_title":"FlowTTS-GRPO: Online Reinforcement Learning with Multi-Objective Reward Optimization for Flow-Matching Based Text-to-Speech","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2606.21343","citing_title":"An Evaluation Framework for Text-to-Speech Voice Reconstruction","ref_index":43,"is_internal_anchor":false},{"citing_arxiv_id":"2606.16417","citing_title":"Joycent: Diffusion-based Accent TTS without Accented Phone Prediction","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2606.11611","citing_title":"SARA: A Dual-Stream VAE for High-Fidelity Speech Generation via Integrating Semantic and Acoustic Representations","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2606.09019","citing_title":"TLDR: Compressing Audio Tokens for Efficient Autoregressive Text-to-Speech","ref_index":44,"is_internal_anchor":false},{"citing_arxiv_id":"2606.20650","citing_title":"EmoInstruct-TTS: Dual-Path Instruction-Guided Emotional Speech Synthesis","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2606.06940","citing_title":"Beyond Semantic Dominance: Cognitive Affective Reasoning and Empathetic Response Alignment in Audio Language Models","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2606.05896","citing_title":"Resonant Minds: Closed-Loop Social Avatars with Theory of Mind","ref_index":55,"is_internal_anchor":false},{"citing_arxiv_id":"2606.01677","citing_title":"UniVocal: Unified Speech-Singing Code-Switching Synthesis","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2606.01016","citing_title":"PolySpeech-100: A Large-Scale Benchmark for Speech Understanding Across 100+ Languages and Dialects","ref_index":107,"is_internal_anchor":false},{"citing_arxiv_id":"2605.30993","citing_title":"SwanVoice: Expressive Long-Form Zero-Shot Speech Synthesis for Both Monologue and Dialogue","ref_index":63,"is_internal_anchor":false},{"citing_arxiv_id":"2606.00066","citing_title":"DUET: Unified Dual-Space Emotion Control for Diffusion and Flow-Matching Driven Text-to-Speech","ref_index":44,"is_internal_anchor":false},{"citing_arxiv_id":"2606.31128","citing_title":"UniSAE: Unified Speech Attribute Editing on Speaker, Emotion and Low-Level Content via Discrete Phonetic Posteriorgram Modelling","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2606.29031","citing_title":"How to Leverage Synthetic Speech for LLM-Based ASR Systems?","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2605.22083","citing_title":"RobustSpeechFlow: Learning Robust Text-to-Speech Trajectories via Augmentation-based Contrastive Flow Matching","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2601.03170","citing_title":"TED-TTS: Training-Free Intra-Utterance Emotion and Duration Control for Text-to-Speech Synthesis","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11866","citing_title":"AuDirector: A Self-Reflective Closed-Loop Framework for Immersive Audio Storytelling","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17583","citing_title":"AgentSteerTTS: A Multi-Agent Closed-Loop Framework for Composite-Instruction Text-to-Speech","ref_index":74,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16026","citing_title":"From Flat Language Labels to Typological Priors: Structured Language Conditioning for Multilingual Speech-to-Speech Translation","ref_index":40,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15202","citing_title":"DeepSlide: From Artifacts to Presentation Delivery","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2601.22143","citing_title":"JUST-DUB-IT: Video Dubbing via Joint Audio-Visual Diffusion","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2604.00688","citing_title":"OmniVoice: Towards Omnilingual Zero-Shot Text-to-Speech with Diffusion Language Models","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2604.01897","citing_title":"FastTurn: Unifying Acoustic and Streaming Semantic Cues for Low-Latency and Robust Turn Detection","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11866","citing_title":"AuDirector: A Self-Reflective Closed-Loop Framework for Immersive Audio Storytelling","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2604.26347","citing_title":"The False Resonance: A Critical Examination of Emotion Embedding Similarity for Speech Generation Evaluation","ref_index":18,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ZI73ZZXFEI2HLXCVFMITQMZM2T","json":"https://pith.science/pith/ZI73ZZXFEI2HLXCVFMITQMZM2T.json","graph_json":"https://pith.science/api/pith-number/ZI73ZZXFEI2HLXCVFMITQMZM2T/graph.json","events_json":"https://pith.science/api/pith-number/ZI73ZZXFEI2HLXCVFMITQMZM2T/events.json","paper":"https://pith.science/paper/ZI73ZZXF"},"agent_actions":{"view_html":"https://pith.science/pith/ZI73ZZXFEI2HLXCVFMITQMZM2T","download_json":"https://pith.science/pith/ZI73ZZXFEI2HLXCVFMITQMZM2T.json","view_paper":"https://pith.science/paper/ZI73ZZXF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.21619&json=true","fetch_graph":"https://pith.science/api/pith-number/ZI73ZZXFEI2HLXCVFMITQMZM2T/graph.json","fetch_events":"https://pith.science/api/pith-number/ZI73ZZXFEI2HLXCVFMITQMZM2T/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZI73ZZXFEI2HLXCVFMITQMZM2T/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZI73ZZXFEI2HLXCVFMITQMZM2T/action/storage_attestation","attest_author":"https://pith.science/pith/ZI73ZZXFEI2HLXCVFMITQMZM2T/action/author_attestation","sign_citation":"https://pith.science/pith/ZI73ZZXFEI2HLXCVFMITQMZM2T/action/citation_signature","submit_replication":"https://pith.science/pith/ZI73ZZXFEI2HLXCVFMITQMZM2T/action/replication_record"}},"created_at":"2026-07-05T12:03:49.475546+00:00","updated_at":"2026-07-05T12:03:49.475546+00:00"}