Devices, methods, and program products are disclosed for detecting multiple utterances. One device includes a component. The component is configured to operate with multiple predetermined utterances. The component is configured to detect a predetermined utterance of the multiple predetermined utterances in an audio input in any operational state of the device. The component is configured to store information indicating successful detections of the predetermined utterance, unsuccessful detections of the predetermined utterance, or a combination thereof. The component is configured to transmit the information while the device is in a regular-power operational state.

Devices, methods, and program products are disclosed for detecting multiple utterances. One device includes a component. The component is configured to operate with multiple predetermined utterances. The component is configured to detect a predetermined utterance of the multiple predetermined utterances in an audio input in any operational state of the device. The component is configured to store information indicating successful detections of the predetermined utterance, unsuccessful detections of the predetermined utterance, or a combination thereof. The component is configured to transmit the information while the device is in a regular-power operational state.

An electronic apparatus includes: a memory storing at least one instruction; and at least one processor configured to divide audio data into a plurality of periods to include overlapping regions, acquire an audio feature from each of the plurality of divided periods, identify a first audio source and a second audio source in each of the plurality of divided periods based on the audio feature, and acquire first audio data corresponding to the first audio source and second audio data corresponding to the second audio source from the audio data.

There is provided a system and method for recognizing tone patterns in spoken languages using sequence-to-sequence neural networks in an electronic device. The recognized tone patterns can be used to improve the accuracy for a speech recognition system on tonal languages.

There is provided a system and method for recognizing tone patterns in spoken languages using sequence-to-sequence neural networks in an electronic device. The recognized tone patterns can be used to improve the accuracy for a speech recognition system on tonal languages.

A method and an electronic device for detecting crosstalk data are provided. The method for detecting crosstalk data can detect whether an audio data stream includes crosstalk data.

A method and an electronic device for detecting crosstalk data are provided. The method for detecting crosstalk data can detect whether an audio data stream includes crosstalk data.

A signal processing apparatus that handles an adaptive enhancement of a speech signal, receives a first signal and a second signal from a determined source. At least one of a speech signal or at least one noise signal is present in the first signal or the second signal. The first signal and the received second signal are processed to obtain a processed signal for amplification of a gain associated with the speech signal present in the first signal and the second signal by a determined factor. A signal-to-noise ratio (SNR) associated with the processed signal is greater than or equal to a threshold value. A reference noise signal is obtained from the second signal based on subtraction of an estimated the speech signal present in the received second signal from the processed signal. A processed speech signal is determined based on filtration of the obtained reference noise signal.

A sub-band processing system for reducing computational complexity and memory requirements is disclosed. The sub-band processing system includes: a first logic that partitions and stores a frequency spectrum of bins of real and imaginary data into a smaller number of sub-bands; a second logic that executes a first lossy compression for a first set of the sub-bands, wherein the first set includes those sub-bands having indices that are greater than or equal to a first index; and a third logic that executes, subsequent to a frequency spectrum processing of the lossy compressed data rendered by the second logic, a second lossy compression for a second set of the sub-bands, wherein the second set includes those sub-bands having indices that are less than the first index and greater than or equal to a second index.

An audio encoder for encoding an audio signal having a lower frequency band and an upper frequency band includes: a detector for detecting a peak spectral region in the upper frequency band of the audio signal; a shaper for shaping the lower frequency band using shaping information for the lower band and for shaping the upper frequency band using at least a portion of the shaping information for the lower band, wherein the shaper is configured to additionally attenuate spectral values in the detected peak spectral region in the upper frequency band; and a quantizer and coder stage for quantizing a shaped lower frequency band and a shaped upper frequency band and for entropy coding quantized spectral values from the shaped lower frequency band and the shaped upper frequency band.