Samples of Adversarial DNN-Based Voice Conversion

This page contains following speech samples: Natural (source): Natural speech of source speaker. Natural (target): Natural speech of target speaker. MGE-FF: Minimum generation error training with Feed-Forward neural networks. MGE-HW: Minimum generation error training with highway networks. ADV-FF: Adversarial training with Feed-Forward neural networks. ADV-HW: Adversarial training with highway net

[masatoi]

SP研究会のポスターセッションで面白かった敵対的DNN音声変換のデモ音声