A microphone unit has a transducer, for generating an electrical audio signal from a received acoustic signal; a speech coder, for obtaining compressed speech data from the audio signal; and a digital output, for supplying digital signals representing said compressed speech data. The speech coder may be a lossy speech coder, and may contain a bank of filters with centre frequencies that are non-uniformly spaced, for example mel frequencies.

A very coarse quantization exceeding the measure determined by the masking threshold without or only very little quality losses is enabled by quantizing not immediately the prefiltered signal, but a prediction error obtained by forward-adaptive prediction of the prefiltered signal. Due to the forward adaptivity, the quantizing error has no negative effect on the prediction on the decoder side.

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for identifying a user in a multi-user environment. One of the methods includes receiving, by a first user device, an audio signal encoding an utterance, obtaining, by the first user device, a first speaker model for a first user of the first user device, obtaining, by the first user device for a second user of a second user device that is co-located with the first user device, a second speaker model for the second user or a second score that indicates a respective likelihood that the utterance was spoken by the second user, and determining, by the first user device, that the utterance was spoken by the first user using (i) the first speaker model and the second speaker model or (ii) the first speaker model and the second score.

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for identifying a user in a multi-user environment. One of the methods includes receiving, by a first user device, an audio signal encoding an utterance, obtaining, by the first user device, a first speaker model for a first user of the first user device, obtaining, by the first user device for a second user of a second user device that is co-located with the first user device, a second speaker model for the second user or a second score that indicates a respective likelihood that the utterance was spoken by the second user, and determining, by the first user device, that the utterance was spoken by the first user using (i) the first speaker model and the second speaker model or (ii) the first speaker model and the second score.

Systems, methods, and devices of a voice-directed inspection system that supports multiple inspectors in the inspection of business assets are described. Inspection plans for large and complex business assets can involve several steps. It is advantageous to split large inspection plans into non-overlapping sections to allow multiple inspectors to perform concurrent inspections. Such sectionalizing is also useful in training new inspectors.

Systems, methods, and devices of a voice-directed inspection system that supports multiple inspectors in the inspection of business assets are described. Inspection plans for large and complex business assets can involve several steps. It is advantageous to split large inspection plans into non-overlapping sections to allow multiple inspectors to perform concurrent inspections. Such sectionalizing is also useful in training new inspectors.

Disclosed examples include accessing sensor data; recognizing, by executing an instruction with programmable circuitry, a feature in the sensor data based on a convolutional neural network; and transitioning, by executing an instruction with the programmable circuitry, a mobile device between at least two of motion feature detection, audio feature detection, or camera feature detection after the feature is recognized in the sensor data, the mobile device to operate at a different level of power consumption after the transition than before the transition.

Disclosed examples include accessing sensor data; recognizing, by executing an instruction with programmable circuitry, a feature in the sensor data based on a convolutional neural network; and transitioning, by executing an instruction with the programmable circuitry, a mobile device between at least two of motion feature detection, audio feature detection, or camera feature detection after the feature is recognized in the sensor data, the mobile device to operate at a different level of power consumption after the transition than before the transition.

The disclosure provides a packet loss recovery method for an audio data packet an electronic device and a storage medium. The method includes: receiving an audio data packet sent by a vehicle-mounted terminal, and identifying a discarded first sampling point set in response to detecting packet loss; obtaining a second sampling point set and a third sampling point set each adjacent to the first sampling point set, in which the second sampling point set is prior to the first sampling point set, the third sampling point set is behind the first sampling point set; and generating target audio data of the first sampling points based on first audio data sampled at the second sampling points and second audio data sampled at the third sampling points, and inserting the target audio data at sampling positions of the first sampling points.

The application relates to HFR (High Frequency Reconstruction/Regeneration) of audio signals. In particular, the application relates to a method and system for performing HFR of audio signals having large variations in energy level across the low frequency range which is used to reconstruct the high frequencies of the audio signal. A system configured to generate a plurality of high frequency subband signals covering a high frequency interval from a plurality of low frequency subband signals is described. The system comprises means for receiving the plurality of low frequency subband signals; means for receiving a set of target energies, each target energy covering a different target interval within the high frequency interval and being indicative of the desired energy of one or more high frequency subband signals lying within the target interval; means for generating the plurality of high frequency subband signals from the plurality of low frequency subband signals and from a plurality of spectral gain coefficients associated with the plurality of low frequency subband signals, respectively; and means for adjusting the energy of the plurality of high frequency subband signals using the set of target energies.