Text to Speech and Voice Cloning Back-end Development Project

Alef company has already set up AWS servers in Singapore location and China Beijing location for its multilingual speech recognition, text to speech, and translation http web services. The Alef APIs offered to our clients are json files through the endpoints located in AWS Singapore and China. Alef company currently plans to upgrade Text to Speech methods to include following two modes. 1) Train new TTS Model for a specific person with 2 hours length recording and transcription. The requirement is to get natural sounding and clear reproduction of the voice of this person for any text input. Fastspeech2 + Hifinet should be the solution of this method unless you have better idea. 2) Voice Cloning tech with a few minutes customer voice input via open source project at [login to view URL] Add finetuning for specific person is also needed. zhrtvc + finetuning should be the solution for this method for Chinese language. You need to develop the program back-end, provide all source codes and datasets in our Github private repository, and finally deploy the application in our company server side as well as in SDK. The support language should include the following 5 languages: English, Chinese Mandarin, French, German and Dutch. Attached PDF file is the background and reference for this project. Language needed: Python, Json, Pytorch, Tensorflow, Dockerfile