Cross-lingual Voice Conversion

I wish I could speak many languages. Wait. Actually I do. But only 4 or 5 languages with limited proficiency. Instead, can I create a voice model that can copy any voice in any language? Possibly! A while ago, me and my colleage Dabi opened a simple voice conversion project. Based on it, I expanded the idea to cross-languages. I found it's very challenging with my limited knowledge. Unfortunately, the results I have for now are not good, but hopefully it will be helpful for some people.

Requirements

• NumPy >= 1.11.1
• TensorFlow >= 1.3
• librosa
• tqdm
• scipy

Data

• Training 1: TIMIT
• Training 2: CMU ARCTIC SLT
• Conversion Sample Files: 50LANGUAGES MP3 audio files

Architecture

• Train 1: MFCCs of TIMIT speakers -> Triphone PPGs
• Train 2: MFCCs of ARTCTIC speaker -> Triphone PPGs -> linear spectrogram
• Convert: MFCCs of Any speakers -> Triphone PPGs -> linear spectrogram -> (Griffin-Lim) -> wav file

(To see what PPGs are, consult this)

Training

• STEP 0. Prepare datasets
• STEP 1. Run python train1.py for phoneme recognition model.
• STEP 2. Run python train2.py for speech synthesis model.

Training Curves

• Training 1

• Training 2

Sample Synthesis

• Run python convert.py and check the generated samples in 50lang-output folder.

Generated Samples

• Check here and compare original speech samples in 16 languages and their converted counterparts.
• Don't expect too much!

References

• L. Sun, S. Kang, K. Li, and H. Meng, “Personalized, cross-lingual TTS using phonetic posteriorgrams,” in Proc. INTERSPEECH, San Francisco, U.S.A., Sep. 2016, pp. 322–326.
• Dabi Ahn & Kyubyong Park, Voice Conversion with Non-Parallel Data. https://github.com/andabi/deep-voice-conversion