Speech processing apparatus and methods

DETAILED DESCRIPTION It is an object of the present invention to provide improved speech processing apparatus and methods.
According to one aspect of the present invention there is provided speech processing apparatus including microphone means arranged to sense speech sounds made by a speaker and derive a first output signal in accordance therewith, the apparatus including sensor means arranged to monitor environmental influences on the speaker of the kind that modify the speech sounds made by the speaker, and processing means arranged to receive the output from the microphone means and from the sensor means, the processing means being arranged to produce a second output signal in accordance with the speech sounds made by the speaker that is compensated at least in part for the environmental influences.
The sensor means may be responsive to vibration or acceleration forces on the speaker.
Alternatively, the sensor means may be responsive to the level of audible noise heard by the speaker.
The speech processing apparatus preferably includes storage means containing information of a reference vocabulary of words, the storage means being arranged to derive a signal by modifying the information in accordance with the output of the sensor means, and the apparatus including comparator means that is arranged to compare the first output signal with the signal derived by the storage means.
The apparatus may be arranged to perform dynamic template adaptation the reference vocabulary in accordance with the output of the sensor means.
The processing means may contain information about how vowel and consonant groupings are influenced by forces on the speaker, the processing means using this information in producing the second output signal.
The comparator may be arranged to perform comparison by pattern matching techniques.
The storage means may be arranged to carry out active word selection from a vocabulary in accordance with mode data.
The storage means may be arranged to perform sub-set selection from the vocabulary in accordance with words previously spoken and recognised