Speaker localization system

Abstract

This invention provides a sound source localization system, comprising a sound direction analysis module, an ultra-wideband module, and a positioning analysis module. The sound direction analysis module includes a microphone array to receive sound indications emitted by a speaker in space, and can determine the speaker's orientation relative to the microphone array based on the sound indications. The ultra-wideband module is used to determine the breathing state and position of multiple people in space, so as to select at least one candidate speaker from these people. The positioning analysis module connects the sound direction analysis module and the ultra-wideband module to receive the orientation and the position of the candidate speaker. It is used to compare the orientation with a first orientation of at least one candidate speaker relative to the microphone array, and select the candidate speaker with the orientation closest to the first orientation as the speaker.

Claims

1. A sound source localization system for locating a speaker emitting a sound indication in a space, the system comprising: a sound location analysis module including a microphone array disposed in the space to receive the sound indication emitted by the speaker, the sound location analysis module being used to determine the location of the sound indication emitting the sound indication relative to the microphone array; and an ultra-wideband (UWB) microphone array. A UWB (Ultra-Wideband) module is installed in the space and connected to the sound location analysis module. The UWB module is used to determine the breathing state and position of multiple people in the space to select at least one candidate speaker from among them. The UWB module determines whether the person corresponding to the first breathing state is the at least one candidate speaker based on whether the first breathing state has a long exhalation interval. A positioning analysis module is connected to the sound location analysis module and the UWB module to receive the location and the position of the at least one candidate speaker. The positioning analysis module is used to compare the location with the first location of the at least one candidate speaker relative to the microphone array and select the candidate speaker corresponding to the first location closest to the location as the speaker.

2. The sound source localization system of claim 1 further includes a command analysis module connected to the sound direction analysis module and the location analysis module. The command analysis module is used to receive and analyze the indication content of the sound indication from the sound direction analysis module, and generate an actuation command to an actuation unit based on the location of the speaker determined by the location analysis module and the indication content.

3. The sound source localization system as described in claim 2, wherein the instruction analysis module is used to determine whether the sound indication contains a voice keyword, and when the sound indication contains the voice keyword, generates a detection instruction to the ultra-wideband module to control the ultra-wideband module to detect the breathing state of the persons.

4. The sound source localization system of claim 2, wherein the ultra-wideband module is further used to determine one of the gestures of the persons respectively, and the instruction analysis module is used to generate the action instruction based on the instruction content of the sound instruction and the gesture of the speaker.

5. A sound source localization system for locating a speaker emitting a sound instruction in a space, the system comprising: a microphone array disposed in the space to receive the sound instruction emitted by the speaker; and an ultra-wideband module disposed in the space and connected to the microphone array, the ultra-wideband module being used to determine a breathing state and a position of each of a plurality of persons in the space, and to determine the speaker and the speaker's position based on the breathing state and the position; wherein, The ultra-wideband module is used to determine whether the person corresponding to the first breathing state is the speaker based on whether the first breathing state has a long exhalation interval and a short inhalation interval preceding the long exhalation interval.

6. The sound source localization system of claim 5 further includes a command analysis module connected to the microphone array and the ultra-wideband module, the command analysis module being used to receive and analyze the indication content of the sound indication from the microphone array, and generate an actuation command to an actuation unit based on the location of the speaker determined by the ultra-wideband module and the indication content.

7. The sound source localization system as claimed in claim 6, wherein the instruction analysis module is used to determine whether the sound indication contains a voice keyword, and when the sound indication contains the voice keyword, generates a detection instruction to the ultra-wideband module to control the ultra-wideband module to detect the breathing state of the persons.

8. The sound source localization system of claim 6, wherein the ultra-wideband module is further used to determine one of the gestures of the persons respectively, and the instruction analysis module is used to generate the action instruction based on the instruction content of the sound instruction and the gesture of the speaker.

9. The sound source localization system of claim 5, wherein the microphone array is used to detect an ambient sound wave reflection value in the space, and when the ambient sound wave reflection value exceeds a threshold, the ultra-wideband module is used to determine the speaker and the speaker's location based on the breathing state and the position.

Images