[INTERSPEECH 2020] VQVC+: One-Shot Voice Conversion by Vector Quantization and U-Net architecture | Hung-yi Lee | Podwise