WaveEx: Accelerating Flow Matching-based Speech Generation via Wavelet-guided Extrapolation

Flow matching-based generative models offer a principled approach to modeling continuous-time dynamics in speech generation. However, inference is often computationally expensive due to repeated neural network evaluations required by ODE solvers. We propose WaveEx, a training-free and plug-in acceleration framework which replaces portions of ODE integration with wavelet-guided extrapolation. By leveraging the multi-scale structure of latent trajectories, WaveEx predicts future states directly in the frequency domain without additional model evaluations or architectural changes. WaveEx consistently accelerates inference across diverse speech generation tasks. The gains are especially pronounced in tasks like speech synthesis (up to 5.73× speedup) and music generation (2.75×), where flow matching plays a central role in alignment modeling and dense ODE integration. Even in tasks with simpler input-output mappings such as speech enhancement (4.55×) and voice conversion (2.75×), WaveEx still achieves notable acceleration, demonstrating the robustness and generalizability of the approach. These results highlight wavelet-guided extrapolation as a lightweight and broadly applicable alternative to full ODE solving for flow matching-based speech generation.

Authors

• Xiaoqian Liu Northeastern University
• Xiyan Gui Huazhong University of Science and Technology Shanghai Jiao Tong University
• Zhengkun Ge Northeastern University
• Yuan Ge Northeastern University
• Chang Zou
• Jiacheng Liu
• Zhikang Niu
• Qixi Zheng Shanghai Jiaotong University
• Chen Xu College of Computer Science and Technology, Harbin Engineering University, China
• Xie Chen Shanghai Jiao Tong University, Shanghai, China
• Tong Xiao School of Computer Science and Engineering, Northeastern University, Shenyang, China NiuTrans Research, Shenyang, China
• Jingbo Zhu School of Computer Science and Engineering, Northeastern University, Shenyang, China NiuTrans Research, Shenyang, China
• Linfeng Zhang Shanghai Jiaotong University

Keywords

No keywords are indexed for this paper.