Speaker and mother tongue independent analysis and recognition of some nasals, liquids and fricatives for integration in an automatic speech recognition system.