- the approch to achieve the multilanguage support depending on language string resource py file:
    - `zh.py` for Chinese
    - `en.py` for english
  - for convenience,these module was managed by the `multilanguage.py`module.
  - unfortunately,these features was test in `fviewer.py`,the main gui was still unable to work normally because the `windows.close` caused the program broken!
- improve the regex_filter performance by using `re.compile` instead `re.<function>`

- update `npy` save strategy

- "there are some problems with recognize the single audio with pca preprocessing,the idea is to save the pca transformer when it was first fited,and then try to load corresponding file(joblib or pickle) when you need to recognize several audios that need extract the required dimension feature"

update the requirements.txt and related introduction in readme;update utils.py to deal with pickel and time field tag str

add the SAVEE speech db,fix some gitches;improve the robustness in visiting files;add some tkinter demos