×

Linear filtering for noise-suppressed speech detection via multiple network microphone devices

  • US 10,692,518 B2
  • Filed: 09/29/2018
  • Issued: 06/23/2020
  • Est. Priority Date: 09/29/2018
  • Status: Active Grant
First Claim
Patent Images

1. A first NMD (“

  • NMD”

    ) comprising;

    a plurality of microphones comprising a first microphone and a second microphone;

    one or more processors;

    a network interface; and

    tangible, non-transitory, computer-readable media storing instructions executable by the one or more processors to cause the first NMD to perform operations comprising;

    receiving an instruction to process one or more audio signals captured by a second NMD comprising a third microphone, wherein the first and second NMDs are separate devices that are positioned at different physical locations within an environment;

    after receiving the instruction,functionally disabling the first microphone of the first NMD,capturing a first audio signal via the second microphone of the first NMD, wherein the first audio signal received at the first NMD comprises first noise content from a noise source, andreceiving over the network interface a second audio signal captured via at least the third microphone of the second NMD, wherein the second audio signal received at the second NMD comprises second noise content from the noise source;

    identifying the first noise content in the first audio signal captured by the first NMD;

    using the identified first noise content from the first NMD to determine an estimated noise content captured by at least the second microphone of the first NMD and the third microphone of the second NMD;

    using the estimated noise content to suppress the first noise content in the first audio signal and the second noise content in the second audio signal;

    generating a composite audio signal by combining the suppressed first audio signal and the suppressed second audio signal;

    determining that the composite audio signal includes a voice input comprising a wake word; and

    in response to the determination, processing the voice input to identify a voice utterance different from the wake word.

View all claims
  • 2 Assignments
Timeline View
Assignment View
    ×
    ×