Provided by: International Journal on Computer Science and Technology (IJCST)
Topic: Enterprise Software
Current speaker recognizer systems the speaker specific source information at different levels. In this, the authors exploit the source information (LP residual) present at different levels namely sub-segmental, segmental and supra-segmental. The sub-segmental analysis considers LP residual in blocks of 5msec with shift of 2.5msec to extract speaker information. The segmental analysis extracts speaker information by processing in blocks of 20msec with shift of 2.5msec. The supra-segmental speaker information is extracted by viewing in blocks of 250msec with shift of 6.25msec.