MOHD NAZID, H. et al. Improved Speaker-Independent Emotion Recognition from Speech Using Two-Stage Feature Reduction. Journal of Information and Communication Technology, v. 14, p. 57–76, 28 Apr.2015.