Fast sequential floating forward selection applied to emotional speech features estimated on

In this paper, we classify speech into several emotional states based on the statistical properties of prosody features estimated on utter-ances extracted from Danish Emotional Speech (DES) and a subset of Speech Under Simulated and Actual Stress (SUSAS) data collec-tions. The proposed novelties are in: 1) speeding up the sequential oating feature selection up to 60%, 2) applying fusion of decisions taken on short speech segments in order to derive a unique decision for longer utterances, and 3) demonstrating that gender and accent information reduce the classication error. Indeed, a lower classi-cation error by 1 % to 11 % is achieved, when the combination of decisions is made on long phrases and an error reduction by 2%-11 % is obtained, when the gender and the accent information is exploited. The total classication error reported on DES is 42.8%. The same gure on SUSAS is 46.3%. The reported human errors have been 32.3 % in DES and 42 % in SUSAS. For comparison pur-poses, a random classication would yield an error of 80 % in DES and 87.5 % in SUSAS, respectively. 1