A text-to-phoneme conversion with backpropagation neural network implemented using an INMOS T800 transputer module (TRAM) parallel processor and phoneme-to-speech conversion using a sound blaster

A text-to-speech synthesis system is presented in this study. The system uses a backpropagation neural network to transform English text into its equivalent phonemes. The output of the neural network is converted to sounds and a sound blaster is used to produce speech from the concatenated phonemes. The system is implemented using parallel processors, INMOS T800 transputer modules (TRAM), mounted on a TBX10/TBX05 motherboard and an Intel 486DX PC serves as the host computer