Portable lip reading sensor system
View Patent ↗The invention comprises a lip reading device having a capacitive array for enhanced portable speech recognition. The capacitive array of the invention produces a sequence of signal frames or signal data sets (i.e., digitized output) representative of the proximity and motion of a user's lips at a predetermined sample rate and resolution. The sequence of signal data sets is stored in a first electronic memory and are compared against a reference data set representative of a predetermined acoustic signal stored in a second electronic memory. The attributes of signal data set are compared against the reference data set for likely data matches based on predetermined criteria.
1. A lip reading sensor system comprising:
a capacitive array comprising a plurality of non-contact proximity-detecting capacitive sensing array pairs having an array output representative of the movement of a user's lips or facial features,
first electronic circuitry for receiving and processing of the array output into a sequence of signal data sets having a predetermined frame rate,
a first electronic memory for the receiving and storing of the sequence of signal data sets,
a second memory for storing at least one reference data set representative of at least one predetermined set of acoustic signals, and,
second electronic circuitry for comparing at least one of the signal data sets with at least one of the references data sets.
2. The sensor system of claim 1 further comprising an acoustic input.
3. The sensor system of claim 1 wherein the capacitive array comprises a 640×480 array of capacitive sensing array pairs.
4. The sensor system of claim 1 wherein the capacitive array has a surface area of about six cm. by about four cm.
5. The sensor system of claim 1 wherein the first electronic circuitry produces a signal data frame rate of at least ten frames per second.
6. A method for enhancing a speech input in a lip reading device comprising:
converting the lip or facial movements of a user using a capacitive array comprising a plurality of non-contact proximity-detecting capacitive sensing array pairs into sequences of signal data sets,
storing at least one of the signal data sets in a first electronic memory,
providing a reference data set representative of a predetermined set of acoustic signals,
comparing at least one of the signal data sets with at least one of the reference data sets for the identification of one or more predetermined attributes, and,
generating an output when one of the predetermined attributes is identified.