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