An electronic device including a memory storing signal information corresponding to a trigger speech; a microphone; a communication interface; and a processor configured to identify whether a first speech signal received through the microphone corresponds to the trigger speech based on the signal information, obtain a first speech sharpness value of the first speech signal based on the identifying, obtain a second speech sharpness value from the at least one external device through the communication interface, based on the first speech sharpness value being greater than the second speech sharpness value, identify a speech command included in the second speech signal received through the microphone by entering a speech recognition mode, and control the electronic device based on the identifying of the speech command, and the processor is further configured to, based on the speech command being unidentified, control the communication interface to transmit a control signal to the at least one external device based on the second speech sharpness value.

A device and method for correcting in-phase and quadrature phase (IQ) baseband components to drive a speaker is provided. The device: controls a local oscillator of an RF downmixing device to a plurality of baseband frequency offsets over a range that includes a given baseband frequency offset; determines, at the plurality of baseband frequency offsets, for a received RF signal, amplitude ratio error and phase error for respective IQ baseband components of the received RF signal; generates, using the amplitude ratio error and the phase error for the respective IQ baseband components, for the given offset, filter coefficients for a given baseband frequency range which compensates for respective amplitude ratio error and respective phase error for the given baseband frequency range; and filters, with the filter coefficients, IQ baseband components of the received RF signal, with the local oscillator operating at the given offset, to generate corrected IQ baseband components.

A device and method for correcting in-phase and quadrature phase (IQ) baseband components to drive a speaker is provided. The device: controls a local oscillator of an RF downmixing device to a plurality of baseband frequency offsets over a range that includes a given baseband frequency offset; determines, at the plurality of baseband frequency offsets, for a received RF signal, amplitude ratio error and phase error for respective IQ baseband components of the received RF signal; generates, using the amplitude ratio error and the phase error for the respective IQ baseband components, for the given offset, filter coefficients for a given baseband frequency range which compensates for respective amplitude ratio error and respective phase error for the given baseband frequency range; and filters, with the filter coefficients, IQ baseband components of the received RF signal, with the local oscillator operating at the given offset, to generate corrected IQ baseband components.

To provide an anomalous sound detection training technique by which a feature amount extraction function for detecting anomalous sound can be generated irrespective of whether training data for anomalous signals is available or not. An anomalous sound detection training apparatus includes: a first function updating unit 3 that updates a feature amount extraction function and an feature amount inverse transformation function, which are input, based on an optimization index of a variational autoencoder; an acoustic feature extraction unit 4 that extracts an acoustic feature of normal sound based on training data for normal sound; a normal sound model updating unit 5 that updates a normal sound model by using the acoustic feature that is extracted; a threshold updating unit 6 that obtains a threshold φ.sub.ρ corresponding to a false positive rate ρ, which has a predetermined value, by using the training data for normal sound and the feature amount extraction function that is input; and a second function updating unit 8 that updates the feature amount extraction function that is updated, based on a Neyman-Pearson-type optimization index defined by the threshold φ.sub.ρ that is obtained, and repeatedly performs processing of each of the above-mentioned units.

To provide an anomalous sound detection training technique by which a feature amount extraction function for detecting anomalous sound can be generated irrespective of whether training data for anomalous signals is available or not. An anomalous sound detection training apparatus includes: a first function updating unit 3 that updates a feature amount extraction function and an feature amount inverse transformation function, which are input, based on an optimization index of a variational autoencoder; an acoustic feature extraction unit 4 that extracts an acoustic feature of normal sound based on training data for normal sound; a normal sound model updating unit 5 that updates a normal sound model by using the acoustic feature that is extracted; a threshold updating unit 6 that obtains a threshold φ.sub.ρ corresponding to a false positive rate ρ, which has a predetermined value, by using the training data for normal sound and the feature amount extraction function that is input; and a second function updating unit 8 that updates the feature amount extraction function that is updated, based on a Neyman-Pearson-type optimization index defined by the threshold φ.sub.ρ that is obtained, and repeatedly performs processing of each of the above-mentioned units.

An examination image storage stores a plurality of examination images having image-capturing time information. A voice processing unit extracts information regarding a finding by recognizing voice that is input to a microphone, and an extracted information storage stores the extracted information regarding the finding and voice time information in association with each other. A grouping processing unit groups a plurality of examination images into one or more image groups based on the image-capturing time information. An association processing unit associates the information regarding the finding stored in the extracted information storage with an image group based on the voice time information. When one examination image is selected by a user, a finding selection screen generation unit generates a screen that displays information regarding a finding associated with an image group including the examination image that is selected.

An examination image storage stores a plurality of examination images having image-capturing time information. A voice processing unit extracts information regarding a finding by recognizing voice that is input to a microphone, and an extracted information storage stores the extracted information regarding the finding and voice time information in association with each other. A grouping processing unit groups a plurality of examination images into one or more image groups based on the image-capturing time information. An association processing unit associates the information regarding the finding stored in the extracted information storage with an image group based on the voice time information. When one examination image is selected by a user, a finding selection screen generation unit generates a screen that displays information regarding a finding associated with an image group including the examination image that is selected.

A method and electronic device for processing a waveform are disclosed. The waveform is representative of bodily sounds. The method includes acquiring the waveform from the sound recording component and having a low-frequency component and a high-frequency component, selecting a target moving averaging filter amongst a first moving averaging filter and a second moving averaging filter for filtering the waveform. The first moving averaging filter is to be used for preserving the low-frequency component of the waveform, and the second moving averaging filter is to be used for preserving the high-frequency component of the waveform. The method includes applying the target moving averaging filter on the waveform for reducing noise in the waveform, thereby generating a second waveform.

A method and electronic device for processing a waveform are disclosed. The waveform is representative of bodily sounds. The method includes acquiring the waveform from the sound recording component and having a low-frequency component and a high-frequency component, selecting a target moving averaging filter amongst a first moving averaging filter and a second moving averaging filter for filtering the waveform. The first moving averaging filter is to be used for preserving the low-frequency component of the waveform, and the second moving averaging filter is to be used for preserving the high-frequency component of the waveform. The method includes applying the target moving averaging filter on the waveform for reducing noise in the waveform, thereby generating a second waveform.

Background noise estimators and methods are disclosed for estimating background noise in an audio signal. Some methods include obtaining at least one parameter associated with an audio signal segment, such as a frame or part of a frame, based on a first linear prediction gain, calculated as a quotient between a residual signal from a 0th-order linear prediction and a residual signal from a 2nd-order linear prediction for the audio signal segment. A second linear prediction gain is calculated as a quotient between a residual signal from a 2nd-order linear prediction and a residual signal from a 16th-order linear prediction for the audio signal segment. Whether the audio signal segment comprises a pause is determined based at least on the obtained at least one parameter; and a background noise estimate is updated based on the audio signal segment when the audio signal segment comprises a pause.