A speech processor for the rehabilitation of deaf people

A method for the rehabilitation of deaf people, based on speech processing and sound compression, is presented. The method implements 3 main functions: the spectral analysis of acquired and sampled words to extract the sound distinctive features, their compression and adaptive shift in the frequency band residual and characteristic of each deaf subject, the reconstruction of the words for the hearing and recognition by the deaf subjects. In correspondence, the processor is made of 3 main parts realising these functions. The processor is provided with a database of words and with a user-friendly interface allowing its interactive use even by people not accustomed with computers. The processor has been experimentally evaluated near the Centre for the Recovery of Hearing and Language (R.E.U.L.) of Genoa, Italy