Audio signal processing method and audio signal processing device
The audio signal processing method in accordance with one embodiment receives an audio signal, obtains a first image, estimates room information based on the obtained first image, sets an acoustic parameter according to the estimated room information, applies sound processing to the audio signal according to the set acoustic parameter, and outputs the audio signal subjected to the sound processing.
1. An audio signal processing method comprising:
obtaining a first image;
estimating room information based on the obtained first image;
setting an acoustic parameter according to the estimated room information;
receiving an audio signal;
obtaining a second image at a timing different from a timing when the first image is obtained;
estimating the room information from the obtained second image;
changing the set acoustic parameter based on the estimated room information estimated from the second image;
applying sound processing to the audio signal according to the changed set acoustic parameter; and
outputting the audio signal subjected to the sound processing.
2. The audio signal processing method according to claim 1 , further comprising changing the acoustic parameter based on the audio signal subjected to the sound processing.
3. The audio signal processing method according to claim 2 , wherein the changing changes the acoustic parameter during a predetermined period.
4. The audio signal processing method according to claim 1 , wherein:
the room information includes space information indicating an open space or a closed space, and
the setting sets the acoustic parameter based on the space information.
5. The audio signal processing method according to claim 1 , wherein:
the room information includes at least one of a size of a room, a shape of a room, a material quality, a numerical quantity of people, a numerical quantity of chairs, or a shape of a desk, and
the setting sets the acoustic parameter according to at least one of the size of the room, the shape of the room, the material quality, the numerical quantity of people, the numerical quantity of chairs, or the shape of the desk.
6. The audio signal processing method according to claim 1 , wherein the sound processing includes at least one of noise reduction, gain adjustment, reverberation removal, or reverberation addition.
7. The audio signal processing method according to claim 1 , wherein:
the room information includes desk information indicating a position of a desk, and
the acoustic parameter is set according to the desk information.
8. The audio signal processing method according to claim 1 , wherein the applying applies the sound processing using a learned model that has learned a relationship between a first audio signal and a second audio signal by machine learning, the second audio signal being obtained by removing noise from the first audio signal.
9. The audio signal processing method according to claim 1 , wherein the estimating estimates the room information using a learned model that has learned a relationship between an input image and the room information by machine learning.
10. An audio signal processing device comprising:
a memory storing instructions; and
a processor configured to implement the instructions to execute a plurality of tasks, including:
an obtaining task that obtains a first image;
an estimating task that estimates room information based on the obtained first image;
a setting task that sets an acoustic parameter according to the estimated room information;
a receiving task that receives an audio signal,
wherein the obtaining task obtains a second image at a timing different from a timing when the first image is obtained,
wherein the estimating task estimates the room information from the obtained second image, and
wherein the setting task changes the set acoustic parameter based on the estimated room information estimated from the second image;
a signal processing task that processes the audio signal according to the changed set acoustic parameter; and
an outputting task that outputs the audio signal subjected to the sound processing.
11. The audio signal processing device according to claim 10 , wherein the signal processing task changes the acoustic parameter based on the audio signal subjected to the sound processing.
12. The audio signal processing device according to claim 11 , wherein the signal processing task changes the acoustic parameter during a predetermined period.
13. The audio signal processing device according to claim 10 , wherein:
the room information includes space information indicating an open space or a closed space, and
the setting task sets the acoustic parameter based on the space information.
14. The audio signal processing device according to claim 10 , wherein:
the room information includes at least one of a size of a room, a shape of a room, a material quality, a numerical quantity of people, a numerical quantity of chairs, or a shape of a desk, and
the setting task sets the acoustic parameter according to at least one of the size of the room, the shape of the room, the material quality, the numerical quantity of people, the numerical quantity of chairs, or the shape of the desk.
15. The audio signal processing device according to claim 10 , wherein the processing performed by the sound processing task includes at least one of noise reduction, gain adjustment, reverberation removal, or reverberation addition.
16. The audio signal processing device according to claim 10 , wherein:
the room information includes desk information indicating a position of a desk, and
the setter sets the acoustic parameter according to the desk information.
17. The audio signal processing device according to claim 10 , wherein the signal processing task processes the audio signal according to the set acoustic parameter using a learned model that has learned a relationship between a first audio signal and a second audio signal by machine learning, the second audio signal being obtained by removing noise from the first audio signal.
18. The audio signal processing device according to claim 10 , wherein the estimating task estimates the room information using a learned model that has learned a relationship between an input image and the room information by machine learning.