Wooseok Han's Projects
Real-time Audio Classification (with Emergency situation)
Recorded BAEKJOON problem by rank
This project is a game written in Python. I used web crawling and socket programming. Think of it as a catch mind using photos.
A project that converts the voice files of old Korean presidents and deep-learns them to recognize their voices.
Use FastSpeech2 and HiFi-GAN to easily perform end-to-end Korean speech synthesis.
PyTorch Implementation of Non-autoregressive Expressive (emotional, conversational) TTS based on FastSpeech2, supporting English, Korean, and your own languages.
Multi-Speaker FastSpeech2 applicable to Korean. Description about train and synthesize in detail.
If you have a wav & transcript, can train HiFi-GAN right now.
Blog
Practice essential libraries such as matplotlib or numpy, and lacking classes.
It is Samsung Card's second Data Analytic contest, which predict multi categories of customer reviews.
Review and study Speech Synthesis codes
Practices speech synthesis
PyTorch Implementation of Meta-StyleSpeech : Multi-Speaker Adaptive Text-to-Speech Generation
Tacotron, Korean, Wavenet-Vocoder, Korean TTS
Learn TensorFlow to try out the voice processing techniques