Detection of unintentional air leaks in a user interface (e.g., mask) of a respiratory therapy system (e.g., a positive air pressure device) is disclosed. One or more sensors (e.g., within a computing device, such as a smartphone) can be moved around relative to the user interface to determine a location and/or intensity of an air leak. The computing device can provide feedback regarding the location and/or intensity of the air leak to facilitate the user locating the air leak, and thus correcting the air leak. In some cases, augmented reality annotations can be overlaid on an image (e.g., live image) of the user wearing the user interface to identify the location of the air leak. The system can automatically detect the type of user interface being used and can provide tailored guidance for reducing the air leaks.

Provided is a robot comprising a driving part, a camera, a plurality of microphones arranged in different directions. The robot further comprises a memory storing instructions, and a processor configured to execute the instructions to identify an originating direction of an audio signal input through the plurality of microphones based on the audio signal being identified as corresponding to a coughing sound, control the camera to capture an image in the originating direction, identify a sterilization area based on a position of a user who does not wear a mask identified in the image, and control the driving part to move the robot to the sterilization area.

The invention provides a sound field optimization method, including steps of building up a first acoustic response model for a controlled area inside a car and building up a second acoustic response model for an uncontrolled area inside the car; building up a first acoustic response difference calculation model based on the first acoustic response model and the second acoustic response model; calculating optimal solution of audio parameters; and controlling each speaker to output audio signals according to corresponding optimal solution of audio parameters. Thus, the vocal area isolation is achieved, so as to provide more extreme acoustic hearing, sound field effect in the controlled area where the user is located inside the car is optimized.

A method of performing automatic speech recognition (ASR) using end-pointing markers generated using accelerometer-based voice activity detector starts with a voice activity detector (VAD) generating an accelerometer VAD output (VADa) based on data output by at least one accelerometer that is included in at least one earbud. The at least one accelerometer to detect vibration of the user's vocal chords. A voice processor detects a speech signal based on acoustic signals from at least one microphone. An end-pointer generates the end-pointing markers based on the VADa output and an ASR engine performs ASR on the speech signal based on the end-pointing markers. Other embodiments are also described.

Embodiments disclosed herein may include a wearable apparatus including a frame having a memory and processor associated therewith. The apparatus may include a camera associated with the frame and in communication with the processor, the camera configured to track an eye of a wearer. The apparatus may also include at least one microphone associated with the frame. The at least one microphone may be configured to receive a directional instruction from the processor. The directional instruction may be based upon an adaptive beamforming analysis performed in response to a detected eye movement from the infrared camera. The apparatus may also include a speaker associated with the frame configured to provide an audio signal received at the at least one microphone to the wearer.

Provided are an audio processing method and apparatus, and a storage medium, which relate to the technical field of artificial intelligence and, in particular, to the speech technical field. The specific implementation solution is as follows. In response to receiving to-be-processed audio, a target sounding direction corresponding to the to-be-processed audio is determined; direction sense reconstruction is performed on the to-be-processed audio according to a direction sense reconstruction filter corresponding to the target sounding direction to obtain target audio; and the target audio is output.

A loudspeaker device includes: a first loudspeaker that outputs a sound; a second loudspeaker that is adjacent to the first loudspeaker in a predetermined direction and outputs a sound in a direction intersecting with a direction in which the first loudspeaker outputs the sound; and a phase switching circuit that switches between a first switching state and a second switching state, the first switching state being a state in which an input sound signal is inputted to both the first loudspeaker and the second loudspeaker, the second switching state being a state in which the input sound signal is inputted to one of the first loudspeaker and the second loudspeaker while an inverted sound signal obtained by inverting a phase of the input sound signal is inputted to the other of the first loudspeaker and the second loudspeaker.

Array microphone systems and methods having adjustable lobe shapes are provided. The lobe shapes of pickup patterns in an array microphone may be adjusted by weighting the audio signals of subsets of the microphone elements that make up the array. The lobe shapes may be adjusted in a direction independent of a steering vector of the lobe. Users may have greater control of lobes which can result in more efficient and optimal coverage of audio sources in environments.

In at least one embodiment, a road alert system is provided. The system includes a loudspeaker array and at least one controller. The loudspeaker array transmits an audio output signal to a vehicle traveling on a road. The at least one controller is programmed to receive a message indicative of a warning for the vehicle for transmission on the audio output signal and to apply equalization parameters to the message to increase in-vehicle speech intelligibility of the audio output signal within the vehicle. The at least one controller is further programmed to transmit the audio output signal via the loudspeaker array in a beamforming mode to minimize noise pollution for objects positioned along the road.

An acoustic chamber includes a loudspeaker, which includes at least two membranes that each reproduce a different frequency band, and a filter that makes it possible to generate a plurality of activation signals from an audio signal source. The activation signals are each applied to an actuator of one of the membranes. The acoustic chamber has an operating range having a variable and controlled directivity, each frequency of which belongs to at least two frequency bands reproduced by the membranes. The acoustic chamber obtains a directivity control signal, and the filter makes it possible to dose, for each frequency of the operating range and depending on the directivity control signal, the contribution of each one of the at least two membranes reproducing the frequency.