ASCLSDMay 30, 2023

Voice Conversion With Just Nearest Neighbors

arXiv:2305.18975v1118 citations
Originality Incremental advance
AI Analysis

This provides a more reproducible and straightforward solution for voice conversion tasks, though it is incremental in approach.

The paper tackles any-to-any voice conversion by proposing kNN-VC, a simple method using nearest neighbors on self-supervised representations, which improves speaker similarity with comparable intelligibility to existing methods.

Any-to-any voice conversion aims to transform source speech into a target voice with just a few examples of the target speaker as a reference. Recent methods produce convincing conversions, but at the cost of increased complexity -- making results difficult to reproduce and build on. Instead, we keep it simple. We propose k-nearest neighbors voice conversion (kNN-VC): a straightforward yet effective method for any-to-any conversion. First, we extract self-supervised representations of the source and reference speech. To convert to the target speaker, we replace each frame of the source representation with its nearest neighbor in the reference. Finally, a pretrained vocoder synthesizes audio from the converted representation. Objective and subjective evaluations show that kNN-VC improves speaker similarity with similar intelligibility scores to existing methods. Code, samples, trained models: https://bshall.github.io/knn-vc

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes