VC_Tacotron

Voice Conversion using Tacotron Model.

Quick Start

1. Download dataset and Preprocess

./download_data.sh
python preprocess.py

2. Training

python train.py

3. Evaluation

./eval.sh

Results

1. Audio samples

clb -> dbl

./samples

2. Alignments

Reference

• Tacotron: Towards End-to-End Speech Synthesis,

• keithito's tacotron implementation

• cmu_arctic dataset