Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

JoyVoice: Long-Context Conditioning for Anthropomorphic Multi-Speaker Conversational Synthesis

About

Large speech generation models are evolving from single-speaker, short sentence synthesis to multi-speaker, long conversation geneartion. Current long-form speech generation models are predominately constrained to dyadic, turn-based interactions. To address this, we introduce JoyVoice, a novel anthropomorphic foundation model designed for flexible, boundary-free synthesis of up to eight speakers. Unlike conventional cascaded systems, JoyVoice employs a unified E2E-Transformer-DiT architecture that utilizes autoregressive hidden representations directly for diffusion inputs, enabling holistic end-to-end optimization. We further propose a MM-Tokenizer operating at a low bitrate of 12.5 Hz, which integrates multitask semantic and MMSE losses to effectively model both semantic and acoustic information. Additionally, the model incorporates robust text front-end processing via large-scale data perturbation. Experiments show that JoyVoice achieves state-of-the-art results in multilingual generation (Chinese, English, Japanese, Korean) and zero-shot voice cloning. JoyVoice achieves top-tier results on both the Seed-TTS-Eval Benchmark and multi-speaker long-form conversational voice cloning tasks, demonstrating superior audio quality and generalization. It achieves significant improvements in prosodic continuity for long-form speech, rhythm richness in multi-speaker conversations, paralinguistic naturalness, besides superior intelligibility. We encourage readers to listen to the demo at https://jea-speech.github.io/JoyVoice

Fan Yu, Tao Wang, You Wu, Lin Zhu, Wei Deng, Weisheng Han, Wenchao Wang, Lin Hu, Xiangyu Liang, Xiaodong He, Yankun Huang, Yu Gu, Yuan Liu, Yuxuan Wang, Zhangyu Xiao, Ziteng Wang, Boya Dong, Feng Dang, Jinming Chen, Jingdong Li, Jun Wang, Yechen Jin, Yuan Zhang, Zhengyan Sheng, Xin Wang• 2025

Related benchmarks

TaskDatasetResultRank
Text-to-SpeechSeed-zh (test)
CER0.97
32
Text-to-SpeechSeed-en (test)
WER1.69
30
Text-to-SpeechSEED (test-hard)
CER5.55
10
Multi-speaker Voice CloneJoyVoice zh-msmt (test)
CER (2spk)1.44
5
Multi-speaker Voice CloneJoyVoice MSMT-eval en (test)
WER (2spk)3.36
5
Showing 5 of 5 rows

Other info

Follow for update