BANG & OLUFSEN A/S

Techniques for dynamic audio objectification are described. Embodiments include providing a first audio snippet from an audio signal to a machine learning model trained based on audio snippets labeled with an audio source and receiving, from the machine learning model, a subset of the first audio snippet that is associated with the audio source. Embodiments include, after playing the reconstituted first audio snippet, receiving a changed configuration relating to the audio source. Embodiments include providing a second audio snippet from the audio signal to the machine learning model and receiving, from the machine learning model, a subset of the second audio snippet that is associated with the audio source. Embodiments include playing a reconstituted second audio snippet based on the subset of the second audio snippet and the changed configuration, wherein an audibly perceptible parameter of the audio source is changed in the reconstituted second audio snippet.

A wearable audio device can include at least one speaker to provide an initial audio output signal; at least one sensor arranged to: sense at least one characteristic of the initial audio output signal, sense noise external to the wearable audio device, and provide at least one sensor signal related to at least one characteristic of the sensed initial audio output signal and at least one characteristic of the sensed noise; at least one input arranged to accept an input related to a desired level of attenuation of the noise; a processor arranged to: receive the at least one sensor signal and the selected desired level of attenuation, select at least one coefficient from a coefficient table based on the at least one sensor signal or the selected desired level of attenuation, and provide an attenuation signal related to the selected at least one coefficient; and at least one filter arranged to receive the attenuation signal from the processor and produce a filtered audio signal based on the attenuation signal, wherein the at least one speaker is configured to provide an attenuated audio output based on the filtered audio signal.

The present disclosure relates to a method for performing inference on input data using a neural network and a processing device employing the aforementioned method. The method comprises the steps of obtaining and storing input data, obtaining parameter data indicating the parameters of the first layer and storing the parameter data in a parameter data storage location and processing the input data using the first layer parameter data, to form first layer output data. The method further comprises storing the first layer output data, obtaining parameter data of the second layer and storing the second layer parameter data by replacing the first layer parameter data with the second layer parameter data, processing the first layer output data using the stored second layer parameter data to form second layer output data; and storing the second layer output data.

A wearable audio device can include at least one speaker to provide an initial audio output signal; at least one sensor arranged to: sense at least one characteristic of the initial audio output signal, sense noise external to the wearable audio device, and provide at least one sensor signal related to at least one characteristic of the sensed initial audio output signal and at least one characteristic of the sensed noise; at least one input arranged to accept an input related to a desired level of attenuation of the noise; a processor arranged to: receive the at least one sensor signal and the selected desired level of attenuation, select at least one coefficient from a coefficient table based on the at least one sensor signal or the selected desired level of attenuation, and provide an attenuation signal related to the selected at least one coefficient; and at least one filter arranged to receive the attenuation signal from the processor and produce a filtered audio signal based on the attenuation signal, wherein the at least one speaker is configured to provide an attenuated audio output based on the filtered audio signal.

Techniques for generating an auditory memory for an auditory event are described. An example technique includes obtaining a first audio content associated with an event in an environment. At least one attribute of the environment is determined, based on evaluating the first audio content. At least one emotional attribute associated with the event in the environment is determined, based on evaluating the first audio content. A second audio content is determined, based at least in part on the at least one attribute of the environment. A third audio content is determined, based at least in part on the at least one emotional attribute. An auditory memory including fourth audio content associated with the event in the environment is generated, based on the first audio content, the second audio content, and the third audio content.

A headset comprising a controller configured to receive wind information comprising information relating to a direction of wind in a predetermined coordinate system, direction information relating to a direction of the headset in the predetermined coordinate system, and an input audio signal and to generate an output audio signal based on the wind information, the direction information and the input audio signal.

A headphone device comprising active noise cancellation means (ANC) having a feedback (FB) signal path, a feed forward (FF) signal path and a sidetone signal path, where the headphone is provided with a control member ( 1, 11 ) that can be displaced or rotated between opposite positions, where the control member ( 1, 11 ) is configured such that the gain of the FF signal path and the gain of the sidetone signal path are controllable by the control member ( 1, 11 ) in such manner that when the control member ( 1, 11 ) is at one extreme position, the gain of the FF signal path is maximum and the gain of the sidetone signal path is minimum and when the control member ( 1,11 ) is at the opposite extreme position, the gain of the FF signal path is minimum and the gain of the sidetone signal path is maximum.

A headset comprising a controller configured to receive wind information comprising information relating to a direction of wind in a predetermined coordinate system, direction information relating to a direction of the headset in the predetermined coordinate system, and an input audio signal and to generate an output audio signal based on the wind information, the direction information and the input audio signal.

A headphone device comprising active noise cancellation means (ANC) having a feedback (FB) signal path, a feed forward (FF) signal path and a sidetone signal path, where the headphone is provided with a control member ( 1, 11 ) that can be displaced or rotated between opposite positions, where the control member ( 1, 11 ) is configured such that the gain of the FF signal path and the gain of the sidetone signal path are controllable by the control member ( 1, 11 ) in such manner that when the control member ( 1, 11 ) is at one extreme position, the gain of the FF signal path is maximum and the gain of the sidetone signal path is minimum and when the control member ( 1,11 ) is at the opposite extreme position, the gain of the FF signal path is minimum and the gain of the sidetone signal path is maximum.

The present disclosure relates to a method for performing inference on input data using a neural network and a processing device employing the aforementioned method. The method comprises the steps of obtaining and storing input data, obtaining parameter data indicating the parameters of the first layer and storing the parameter data in a parameter data storage location and processing the input data using the first layer parameter data, to form first layer output data. The method further comprises storing the first layer output data, obtaining parameter data of the second layer and storing the second layer parameter data by replacing the first layer parameter data with the second layer parameter data, processing the first layer output data using the stored second layer parameter data to form second layer output data; and storing the second layer output data.