SIGNATURE WORK
CONFERENCE & EXHIBITION 2024

The paper shows a comprehensive analysis on voice conversion (VC). The paper bridges the gap between artificial intelligence and digital signal processing in voice conversion, mostly a focus of transformation of acoustic properties from one speaker to another speaker while pre‐ serving the linguistic contents. Later, the paper explains the definitions, methodologies of the research has done so far, historical context, and the impact of deep learning approach. The paper highlights the importance of spectral and prosodic feature manipulation, improvement with WaveNet vocoders, addition of sparse representations, and vector quantization mutual information for improving speech quality with high precision. The paper provides potential challenges and future research guidance. This will highlights the importance of deep learning and statistical machine learning methods to improve voice quality and uniqueness. Through detailed information, it will lead to provide profound insights into achieving realistic and per‐ sonalized voice synthesis. The development of personalized voice synthesis marks a break‐ through toward the artificial applications.