Voice/tone conversion with OpenVoice v2 via ONNX Runtime, using the Hinotsuba/OpenVoice-ONNX-v2 models. Reference (target) voices and source clips are pre-extracted from speech datasets in the ghananlpcommunity org. Datasets used as reference voices are not reused as source clips.
1 · Source audio (the voice to re-tone)
Bundled source clip
Pre-extracted from ghananlpcommunity datasets (never the same datasets used as reference).