Texas Instruments Corporation

National Semiconductor Corporation

Wei Ma

Zhenyong Zhang

Ahmad Reza Bahai

Yunhong Li

Lin Sun

Kambiz Homayounfar

Bozhao Tan

Mark Kuhns

Mikko Topi Loikkanen

K Y Martin Lee

Martin Lee

Peter R Holloway

Jian-Yi Wu

Masood Yousefi

Shougang Wang

Peter Kamp

Ian Atkinson

A sound recognition system including time-dependent analog filtered feature extraction and sequencing. An analog front end (AFE) in the system receives input analog signals, such as signals representing an audio input to a microphone. Features in the input signal are extracted, by measuring such attributes as zero crossing events and total energy in filtered versions of the signal with different frequency characteristics at different times during the audio event. In one embodiment, a tunable analog filter is controlled to change its frequency characteristics at different times during the event. In another embodiment, multiple analog filters with different filter characteristics filter the input signal in parallel, and signal features are extracted from each filtered signal; a multiplexer selects the desired features at different times during the event.

Time-based frequency tuning of analog-to-information feature extraction

A duty-cycled acoustic sensor saves power, for example, by operating for relatively short periods of time in a repetitive manner. A sensor bias current provides operating power to the sensor. An output analog signal from the sensor carries the information induced by the sensor upon the bias signal. Capacitive coupling is employed to remove direct DC voltage from the output analog signal to generate an analog input signal for acoustic analysis. A capacitor for capacitive coupling is pre-charged to reduce the charging time of the capacitor as the sensor is being powered up. After the capacitor is sufficiently precharged, acoustic analysis is performed on the analog input signal. The sensor is powered down by substantially blocking current flow through the sensor, which saves power. Results of the acoustic analysis can be used, for example, to control parameters of the duty-cycling of the acoustic sensor.

Duty-cycling microphone/sensor for acoustic analysis

A low power sound recognition sensor is configured to receive an analog signal that may contain a signature sound. Sparse sound parameter information is extracted from the analog signal and compared to a sound parameter reference stored locally with the sound recognition sensor to detect when the signature sound is received in the analog signal. A portion of the sparse sound parameter information is differential zero crossing (ZC) counts. Differential ZC rate may be determined by measuring a number of times the analog signal crosses a threshold value during each of a sequence of time frames to form a sequence of ZC counts and taking a difference between selected pairs of ZC counts to form a sequence of differential ZC counts.

Robust feature extraction using differential zero-crossing counts

A low power sound recognition sensor is configured to receive an analog signal that may contain a signature sound. Sound parameter information is extracted from the analog signal and compared to a sound parameter reference stored locally with the sound recognition sensor to detect when the signature sound is received in the analog signal. A trigger signal is generated when a signature sound is detected. A portion of the extracted sound parameter information is sent to a remote training location for adaptive training when a signature sound detection error occurs. An updated sound parameter reference from the remote training location is received in response to the adaptive training.

Cloud based adaptive learning for distributed sensors

A frequency selectable driver circuit (driver circuit) includes a current drive transistor between a first and second node for driving a frequency selectable oscillator (FSO) coupled therebetween that includes a first LC network including an inductor Land a capacitor Cin series and a second LC network in series with the first LC network. The second LC network includes a capacitor bank including capacitors in parallel including switches for switching (selectable capacitors) and an inductor Lin parallel to the capacitor bank. Values of Land Cprovide a low frequency point fwhich the LC network is overall capacitive, and in a frequency range between fand fabove fwith only a portion of selectable capacitors selected a capacitance of the capacitor bank and value of Lresults in an overall impedance for the second LC network that is capacitive between fand fand inductive elsewhere.

Ultrasound lens cleaner driver with frequency selectable oscillator

Comparing differential ZC count to database to detect expected sound

A low power sound recognition sensor is configured to receive an analog signal that may contain a signature sound. The received analog signal is evaluated using a detection portion of the analog section to determine when background noise on the analog signal is exceeded. A feature extraction portion of the analog section is triggered to extract sparse sound parameter information from the analog signal when the background noise is exceeded. An initial truncated portion of the sound parameter information is compared to a truncated sound parameter database stored locally with the sound recognition sensor to detect when there is a likelihood that the expected sound is being received in the analog signal. A trigger signal is generated to trigger classification logic when the likelihood that the expected sound is being received exceeds a threshold value.

Powering-up AFE and microcontroller after comparing analog and truncated sounds

A low power sound recognition sensor is configured to receive an analog signal that may contain a signature sound. Sparse sound parameter information is extracted from the analog signal. The extracted sound parameter information is sampled in a periodic manner and a context value is updated to indicate a current environmental condition. The sparse sound parameter information is compared to both the context value and a signature sound parameter database stored locally with the sound recognition sensor to identify sounds or speech contained in the analog signal, such that identification of sound or speech is adaptive to the current environmental condition.

Adaptive environmental context sample and update for comparing speech recognition

A system and method for processing close talking differential microphone array (CTDMA) signals in which incoming microphone signals are transformed from time domain signals to frequency domain signals having separable magnitude and phase information. Processing of the frequency domain signals is performed using the magnitude information, following which phase information is reintroduced using phase information of one of the original frequency domain signals. As a result, high pass filtering effects of conventional differential signal processing of CTDMA signals are substantially avoided.

Frequency domain signal processor for close talking differential microphone array

A system includes multiple speakers arranged in a speaker array configuration. The system also includes one or more filters configured to filter audio signals and generate filtered audio signals. The one or more filters are configured to operate using filter coefficients associated with a desired beam pattern to be produced by the multiple speakers. The system further includes at least one amplifier configured to amplify the filtered audio signals and provide the amplified filtered audio signals to the speakers. The one or more filters reside within or are coupled to the at least one amplifier. The system may further include a controller configured to modify at least one of the filter coefficients based on a change in the speaker configuration. The filters may operate independently of a centralized processor, and a centralized processor may not even be required to provide electronic beam forming.

Growing community of inventors

San Ramon, CA, United States of America

Wei Ma