Abstract: Voice morphing techniques have gained significant attention due to their applications in speech processing and multimedia systems. This study presents a wavelet-based approach for voice morphing, which enables the transformation of a source speaker's voice to resemble a target speaker while maintaining intelligibility and naturalness. The proposed method utilizes the wavelet transform to decompose the source and target speech signals into different frequency subbands. By modifying the wavelet coefficients in these subbands, the spectral and temporal characteristics of the source speech are altered to match those of the target speech. The morphed speech is then synthesized by performing an inverse wavelet transform on the modified coefficients. Objective and subjective evaluations demonstrate the effectiveness of the proposed approach in achieving accurate voice morphing while preserving the linguistic content. The waveletbased voice morphing technique presented in this study offers potential applications in speech synthesis, voice conversion, and entertainment systems.

Keywords: Voice Morphing, Speech Processing, Wavelets, Signal Decomposition, Voice Conversion

PDF | DOI: 10.17148/IJARCCE.2023.12673

Open chat
Chat with IJARCCE