Improved speech recognition using a subspace projection approach

Two class-separability criteria based on the divergence measure are proposed to improve speech recognition performance. The average and weighted average divergence measures are used as criteria for nding a transformation matrix which maps the original features into a more discriminative subspace. Results are presented for a highly confusable task. Address correspondence to: Philipos C. Loizou Department of Applied Science Univ. of Arkansas at Little Rock Little Rock, AR 72204-1099 Phone:(501) 569-8067 Fax:(501) 569-8020 E-mail: [email protected]