Systems and methods are described for generating a transcript of a legal proceeding or other multi-speaker conversation or performance in real time or near-real time using multi-channel audio capture. Different speakers or participants in a conversation may each be assigned a separate microphone that is placed in proximity to the given speaker, where each audio channel includes audio captured by a different microphone. Filters may be applied to isolate each channel to include speech utterances of a different speaker, and these filtered channels of audio data may then be processed in parallel to generate speech-to-text results that are interleaved to form a generated transcript.

Systems and methods are described for generating a transcript of a legal proceeding or other multi-speaker conversation or performance in real time or near-real time using multi-channel audio capture. Different speakers or participants in a conversation may each be assigned a separate microphone that is placed in proximity to the given speaker, where each audio channel includes audio captured by a different microphone. Filters may be applied to isolate each channel to include speech utterances of a different speaker, and these filtered channels of audio data may then be processed in parallel to generate speech-to-text results that are interleaved to form a generated transcript.

An audio system includes: a speaker; a microphone that generates a microphone signal based on sound output from the speaker; a mixer module configured to generate a mixed signal by mixing the microphone signal with an audio signal; a filter module configured to filter the mixed signal to produce a filtered signal and to apply the filtered signal to the speaker; and a detector module configured to determine a howling frequency in the microphone signal attributable to sound output from the speaker, where the filter module is configured to decrease a magnitude of the filtered signal at the howling frequency.

An audio system includes: a speaker; a microphone that generates a microphone signal based on sound output from the speaker; a mixer module configured to generate a mixed signal by mixing the microphone signal with an audio signal; a filter module configured to filter the mixed signal to produce a filtered signal and to apply the filtered signal to the speaker; and a detector module configured to determine a howling frequency in the microphone signal attributable to sound output from the speaker, where the filter module is configured to decrease a magnitude of the filtered signal at the howling frequency.

Techniques for dynamically adjusting received audio are described. In an example, a computer system receives audio data representing noise and utterance received by a device during a first time interval that has a start and an end. The start corresponds to a beginning of the utterance. The end corresponds to at a selection by the device of an audio beam associated with a direction towards an utterance source. The computer system determines a value associated with an audio adjustment factor. The audio adjustment factor is represented by values that vary during the time interval. The value is one of the values associated with a time point of the first time interval. The computer system generates, based at least in part on the audio data and the value, first data that indicates a measurement of at least one of the noise or the utterance.

Techniques for dynamically adjusting received audio are described. In an example, a computer system receives audio data representing noise and utterance received by a device during a first time interval that has a start and an end. The start corresponds to a beginning of the utterance. The end corresponds to at a selection by the device of an audio beam associated with a direction towards an utterance source. The computer system determines a value associated with an audio adjustment factor. The audio adjustment factor is represented by values that vary during the time interval. The value is one of the values associated with a time point of the first time interval. The computer system generates, based at least in part on the audio data and the value, first data that indicates a measurement of at least one of the noise or the utterance.

An electronic device is provided. The electronic device includes, for the purpose of determining a customized beamformer filter, an input module including a plurality of microphones configured to receive an external sound signal, a memory configured to store computer-executable instructions and an initial value of a voice parameter used to perform beamforming on the external sound signal, and a processor configured to execute the instructions by accessing the memory. The instructions may be configured to estimate a feature value of the external sound signal, calculate the initial value of the voice parameter used to perform beamforming based on the external sound signal received by the plurality of microphones, determine whether to store the calculated initial value according to the feature value, determine which one of the calculated initial value or an initial value stored in the memory used according to the feature value, and obtain a target voice parameter.

An electronic device is provided. The electronic device includes, for the purpose of determining a customized beamformer filter, an input module including a plurality of microphones configured to receive an external sound signal, a memory configured to store computer-executable instructions and an initial value of a voice parameter used to perform beamforming on the external sound signal, and a processor configured to execute the instructions by accessing the memory. The instructions may be configured to estimate a feature value of the external sound signal, calculate the initial value of the voice parameter used to perform beamforming based on the external sound signal received by the plurality of microphones, determine whether to store the calculated initial value according to the feature value, determine which one of the calculated initial value or an initial value stored in the memory used according to the feature value, and obtain a target voice parameter.

The present disclosure describes aspects of adaptive energy limiting for transient noise suppression. In some aspects, an adaptive energy limiter sets a limiter ceiling for an audio signal to full scale and receives a portion of the audio signal. For the portion of the audio signal, the adaptive energy limiter determines a maximum amplitude and evaluates the portion with a neural network to provide a voice likelihood estimate. Based on the maximum amplitude and the voice likelihood estimate, the adaptive energy limiter determines that the portion of the audio signal includes noise. In response to determining that the portion of the audio signal includes noise, the adaptive energy limiter decreases the limiter ceiling and provides the limiter ceiling to a limiter module effective to limit an amount of energy of the audio signal. This may be effective to prevent audio signals from carrying full energy transient noise into conference audio.

The present disclosure describes aspects of adaptive energy limiting for transient noise suppression. In some aspects, an adaptive energy limiter sets a limiter ceiling for an audio signal to full scale and receives a portion of the audio signal. For the portion of the audio signal, the adaptive energy limiter determines a maximum amplitude and evaluates the portion with a neural network to provide a voice likelihood estimate. Based on the maximum amplitude and the voice likelihood estimate, the adaptive energy limiter determines that the portion of the audio signal includes noise. In response to determining that the portion of the audio signal includes noise, the adaptive energy limiter decreases the limiter ceiling and provides the limiter ceiling to a limiter module effective to limit an amount of energy of the audio signal. This may be effective to prevent audio signals from carrying full energy transient noise into conference audio.