Monitoring system with speech recognition
First Claim
Patent Images
1. A system comprising:
- at least one sensor of a physiological condition of an individual;
a plurality of audio input transducers distributed throughout a residence of the individual;
at least one discrete, stationary background sound source within the residence with a first transducer of the plurality of transducers located at an output of the at least one discrete, stationary background sound source;
control circuitry coupled to the sensor and the transducers, the control circuitry has a communications port, the control circuitry responds to speech received from the transducers and carries out a speech recognition process thereon, and in accordance with the results thereof, initiates communications with a displaced location via the communications port; and
software executed by the control circuitry, the software compensates each of the plurality of input transducers to cancel background sounds from the at least one discrete, stationary background sound source and to null an acoustic signal from each of the plurality of input transducers based upon differences in phase and amplitude of the background sounds detected among each of the plurality of input transducers, the software executed by the control circuitry carrying out speech recognition of audio provided by the individual from the nulled acoustic signal provided by at least some of the plurality of input transducers, said nulling including detecting an acoustic signal from the at least one discrete, stationary background sound source through the first transducer and from each of at least two second transducers of the plurality of transducers, determining a phase difference between the acoustic signal detected by the first transducer and each of the at least two second transducers, adjusting a phase of the signal detected by the first transducer for each of the at least two second transducers based upon the phase difference of each of the respective second transducers, inverting each of the phase adjusted signals, adjusting an amplitude of the respective inverted signals and adding the respective phase and amplitude adjusted signals to the respective acoustic signal detected by the at least two second transducers to null the at least one discrete, stationary background sound source within the respective acoustic signals from the at least two second transducers.
1 Assignment
0 Petitions
Accused Products
Abstract
A system for monitoring conditions associated with an individual in a region includes at least one speech input transducer and speech processing software coupled thereto. Results of the speech processing can initiate communications with a displaced communications device such as a telephone or a computer to provide a source of feedback.
42 Citations
40 Claims
-
1. A system comprising:
-
at least one sensor of a physiological condition of an individual; a plurality of audio input transducers distributed throughout a residence of the individual; at least one discrete, stationary background sound source within the residence with a first transducer of the plurality of transducers located at an output of the at least one discrete, stationary background sound source; control circuitry coupled to the sensor and the transducers, the control circuitry has a communications port, the control circuitry responds to speech received from the transducers and carries out a speech recognition process thereon, and in accordance with the results thereof, initiates communications with a displaced location via the communications port; and software executed by the control circuitry, the software compensates each of the plurality of input transducers to cancel background sounds from the at least one discrete, stationary background sound source and to null an acoustic signal from each of the plurality of input transducers based upon differences in phase and amplitude of the background sounds detected among each of the plurality of input transducers, the software executed by the control circuitry carrying out speech recognition of audio provided by the individual from the nulled acoustic signal provided by at least some of the plurality of input transducers, said nulling including detecting an acoustic signal from the at least one discrete, stationary background sound source through the first transducer and from each of at least two second transducers of the plurality of transducers, determining a phase difference between the acoustic signal detected by the first transducer and each of the at least two second transducers, adjusting a phase of the signal detected by the first transducer for each of the at least two second transducers based upon the phase difference of each of the respective second transducers, inverting each of the phase adjusted signals, adjusting an amplitude of the respective inverted signals and adding the respective phase and amplitude adjusted signals to the respective acoustic signal detected by the at least two second transducers to null the at least one discrete, stationary background sound source within the respective acoustic signals from the at least two second transducers. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 31, 32)
-
-
10. A monitoring system comprising:
-
at least a first sensor for monitoring a physiological condition of a person and associated sensor interface circuitry; a plurality of acoustical input transducers for monitoring acoustical signals and associated transducer interface circuitry distributed throughout a residence of the person; at least one discrete, stationary background noise source within the residence with a first transducer of the plurality of acoustic input transducers located at an output of the at least one discrete, stationary background noise source; a device that receives at least first sensor data and which receives data from the transducers and respective interface circuitry, to carry out at least a speech recognition function, and responsive thereto, to initiate communications with a displaced unit; and control software executed by the device, the software compensates each of the plurality of input transducers to cancel background sounds from the at least one discrete, stationary sound source and to null an acoustic signal from each of the plurality of input transducers based upon differences in phase and amplitude of the background sounds detected between the first input transducer and each of at least two other input transducers of the plurality of input transducers said canceling including inverting a signal received through the first input transducer of the plurality of input transducers for each of the at least two other input transducers, adjusting an amplitude and phase of the inverted signals based upon the respective detected difference in phase and amplitude and adding the respective inverted and adjusted signals to a signal from the respective at least two input transducers the software executed by the device carrying out speech recognition on audio provided by the person from the nulled acoustic signal provided by the at least two input transducers. - View Dependent Claims (11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30)
-
-
33. A method comprising:
-
providing a plurality of audio input transducers located in an area to be monitored; providing at least one discrete, stationary sound source within the area with a first of the plurality of audio input transducers located at an output of the at least one discrete, stationary sound source; detecting audio received from the area; nulling simultaneous ambient background audio from the at least one discrete, stationary sound source within the detected audio of an acoustic signal from at least two other of the plurality of audio input transducers by detecting a difference in phase and amplitude between a signal from the first transducer and a respective signal from at least two other input transducers of the plurality of transducers, inverting a signal received through the first transducer, adjusting an amplitude and phase of the inverted signal based upon the detected differences in phase and amplitude for each of the at least two other input transducers and adding the respective inverted and adjusted signal to a respective signal from the at least two other transducers; locally performing speech recognition on the detected audio with nulled ambient background audio from at least some of the plurality of input transducers to distinguish a resident'"'"'s speech from simultaneous ambient local background audio from the at least one discrete, stationary sound source; selecting at least one speech output in response to the results of the speech recognition; coupling speech signals from the at least some audio input transducer to a communications port; transmitting a signal from the communications port to at least one remote location, the signal conveying information concerning the resident. - View Dependent Claims (34, 35, 36, 37, 38, 39, 40)
-
Specification