Speech auto tagging
identify the language's attribute(such as language type,gender,accent etc.) use keras. now only implementation language type classify.
1. data prepare
use librispeech and THCHS30 as english and chinese speech corpus
run in shell: python data_prepare
2. train
run: python train.py
3. test
run : python test.py