Voice Conversion samples

These are the audio samples I showed during my PhD thesis defense at UPC-TALP. It contains some utterances converted via Weighted Frequency Warping (WFW) using both parallel training and non-parallel training based on an Iterative Nearest-neighbour plus Conversion Alignment algorithm (INCA). Click here if you want to try the code. By the way, the thesis was recently awarded by RTTH and ISCA SIG-IL!!

Natural speech

Synthetic speech

[Back]