VoxMorph: Scalable Zero-shot Voice Identity Morphing via Disentangled Embeddings

TL;DR

VoxMorph introduces a scalable, zero-shot voice morphing framework that disentangles vocal features for high-fidelity, fine-grained identity manipulation, significantly advancing biometric security and voice synthesis technology.

Contribution

It presents a novel zero-shot voice morphing method that does not require model retraining and enables detailed control over vocal traits using disentangled embeddings.

Findings

Achieves 2.6x improvement in audio quality

Reduces intelligibility errors by 73%

Attains 67.8% attack success rate on speaker verification

Abstract

Morphing techniques generate artificial biometric samples that combine features from multiple individuals, allowing each contributor to be verified against a single enrolled template. While extensively studied in face recognition, this vulnerability remains largely unexplored in voice biometrics. Prior work on voice morphing is computationally expensive, non-scalable, and limited to acoustically similar identity pairs, constraining practical deployment. Moreover, existing sound-morphing methods target audio textures, music, or environmental sounds and are not transferable to voice identity manipulation. We propose VoxMorph, a zero-shot framework that produces high-fidelity voice morphs from as little as five seconds of audio per subject without model retraining. Our method disentangles vocal traits into prosody and timbre embeddings, enabling fine-grained interpolation of speaking style and identity. These embeddings are fused via Spherical Linear Interpolation (Slerp) and synthesized using an autoregressive language model coupled with a Conditional Flow Matching network. VoxMorph achieves state-of-the-art performance, delivering a 2.6x gain in audio quality, a 73% reduction in intelligibility errors, and a 67.8% morphing attack success rate on automated speaker verification systems under strict security thresholds. This work establishes a practical and scalable paradigm for voice morphing with significant implications for biometric security. The code and dataset are available on our project page: https://vcbsl.github.io/VoxMorph/