a. Is for whom?
Vision is an important role in face to face communication besides hearing. During speaking, lips and tongue will move corresponding to spelling. Thus, the pronounced word can be inferred by the position of lips and tongue in some aspect. To deeply auditory impaired people, lip reading is the dominant mode of speech perception since available hearing aids are of little or no help to them.
But lip reading still has some hazards. There are three main problems for deaf lip-readers to identify words. I.Vocal fold activity is invisible. We are not able to distinct voice or voiceless. II.Identify supra-segmental (prosodic) features, such as rhythm, stress pattern, and intonation, is impossible. III.Some consonants appear to be produced at approximately the same place in the mouth but differ only in their manner of production, for example: /t/, /n/,/s/ Because of these hazards, aiding solutions using other sensory function were developed. Vision lip-reading aid is one of these solution, and it is a harmless and very portable one.
The vision lip-reading aid can identify vowel-consonant-vowel (VCV) nonsense syllables that cannot be read from lips and tongue position, and shows hinting patterns of syllable feature on an LED display. Thus it can help lip-readers to recognize what others are saying, and whether speaking or not.
The block diagram of lip-reading aid is in fig.1. It has a microphone to receive sounds. Than the sound signal will be passed to a feature extractor, formed by filters and amplifiers, to get the characteristic of the sound and pass it by voltage signal. The stimulus encoder will identify these signals as features of speech: voicing, voice onset time, intonation patterns, formant frequencies, and gross high frequency energy. Some experiments for finding high discriminability hint patterns are processed by a signal synthesizer controlled by an HP 9817 computer....