A novel approach for Marathi numeral recognition using Bark scale and discrete sine transform method

Marathi is spoken by the native people of Maharashtra. Spoken word recognition in Marathi is widely studied area of research. This paper describes a method for recognition of Marathi Numerals from `Shunya'(zero) to `Nau' (nine) using Bark scale and Discrete Sine Transform. Features extracted using Bark scale are transformed and reduced using statistical properties. A unique method for feature vector reduction is implemented in the proposed work. Speech signal is processed using Discrete Sine Transform which is suitable for reconstruction of speech signal. The comparative analysis of the Bark scale and Discrete Sine Transform method is done with Mel Frequency Cepstral coefficients and Discrete Cosine Transform. Neural network is used for classification.