-
公开(公告)号:US11451922B1
公开(公告)日:2022-09-20
申请号:US16902102
申请日:2020-06-15
Applicant: Amazon Technologies, Inc.
Inventor: Robert Wadia Moses , Jan Aage Abildgaard Pedersen
Abstract: The described implementations of a head-mounted speaker array apparatus (“HMSAA”) include speaker arrays having one or more speakers that, when the HMSAA is worn on the head of a listener, are positioned forward of the pinna of each ear of the listener. For example, the HMSAA may include a first speaker array that is coupled to a frame that, when worn by a listener, positions the first speaker array forward of a first pinna of a first ear of the listener. A second speaker array may also be coupled to the frame at a position that, when the frame is worn by the listener, the second speaker array is positioned forward of a second pinna of the second ear of the listener.
-
公开(公告)号:US10959018B1
公开(公告)日:2021-03-23
申请号:US16252066
申请日:2019-01-18
Applicant: Amazon Technologies, Inc.
Inventor: Guangji Shi , Jan Aage Abildgaard Pedersen , Trausti Thor Kristjansson , Philip Ryan Hilmes
Abstract: A system that performs wall detection, range estimation, and/or corner detection to determine a position of a device relative to acoustically reflective surfaces. The device generates output audio using a loudspeaker, generates microphone audio data using a microphone array, performs beamforming to generate directional audio data and then generates impulse response data for each of a plurality of directions. The device may detect a peak in the impulse response data and determine a distance and/or direction to a reflective surface based on the peak. Based on a number of reflected surfaces and/or direction(s) of the reflected surfaces detected by the device, the device may classify the different directions and estimate where it is in the room, such as whether the device is in a corner, along one wall, or in an open area. By knowing its position relative to the room surfaces, the device may improve sound equalization.
-
公开(公告)号:US10777214B1
公开(公告)日:2020-09-15
申请号:US16456247
申请日:2019-06-28
Applicant: Amazon Technologies, Inc.
Inventor: Guangji Shi , Trausti Thor Kristjansson , Jan Aage Abildgaard Pedersen , Philip Ryan Hilmes
Abstract: A system that performs wall detection, range estimation, and/or corner detection to determine a position of a device relative to acoustically reflective surfaces. The device generates output audio using loudspeaker(s), generates microphone audio data using a microphone array, and generates impulse response data for each of the microphones. The device may generate the impulse response data using an acoustic echo cancellation (AEC) component or multi-channel AEC (MC-AEC). The device may detect a peak in the impulse response data and determine a distance to a reflective surface based on the peak. Based on a number of reflected surfaces detected by the device, the device may classify a position of the device within the room, such as whether the device is in a corner, along one wall, or in an open area. By knowing the position relative to the room surfaces, the device may improve sound equalization and other processing.
-
-