Speech processing can be used to convert texts into speech, speech into speech, or speech into texts. The basis for speech processing is speech recognition, which identifies spoken words and texts before speech processing and divides them into speech elements. These can be phonemes or the smallest tonal distinguishing features. The conversion into texts, spoken responses or instructions is carried out by corresponding speech processing programs. In speech output, small sound units are combined to form a speech stream.
Examples of speech processing are part-of-speech( PoS), speech-to-text( STT) for spoken text input into writing programs, speech-to-speech for simultaneous translation and use in call centers with interactive voice response( IVR), and text-to-speech ( TTS), which is used in the unified messaging service ( UMS), among other applications.