A comparison of several speech parameters for speaker independent speech recognition and speaker recognition