US 12,462,809 B2
Voice registration device, control method, program, and storage medium
Koji Okabe, Tokyo (JP); and Takafumi Koshinaka, Tokyo (JP)
Assigned to NEC CORPORATION, Tokyo (JP)
Appl. No. 18/016,571
Filed by NEC Corporation, Tokyo (JP)
PCT Filed Jul. 27, 2020, PCT No. PCT/JP2020/028724
§ 371(c)(1), (2) Date Jan. 17, 2023,
PCT Pub. No. WO2022/024188, PCT Pub. Date Feb. 3, 2022.
Prior Publication US 2023/0282217 A1, Sep. 7, 2023
Int. Cl. G10L 17/04 (2013.01); G10L 17/02 (2013.01); G10L 17/20 (2013.01); G10L 21/0208 (2013.01)
CPC G10L 17/04 (2013.01) [G10L 17/02 (2013.01); G10L 17/20 (2013.01); G10L 21/0208 (2013.01)] 9 Claims
OG exemplary drawing
 
1. A voice registration device comprising:
at least one memory storing instructions; and
at least one processor configured to execute the instructions to perform processing comprising:
providing instructions to start reproduction of noise;
in response to providing of the instructions to start the reproduction of the noise, starting the reproduction of the noise;
after starting the reproduction of the noise, providing instructions to start voice input;
in response to the instructions to start the voice input, starting the voice input and thereafter ending the voice input;
after ending of the voice input, providing instructions to end the reproduction of the noise;
in response to the instructions to end the reproduction of the noise, ending the reproduction of the noise;
after the ending of the voice input, determining whether or not to reacquire the voice input, based on a comparison between silent environment voice input and the voice input,
in response to determining to reacquire the voice input, repeating the processing with providing of the instructions to start the reproduction of the noise such that the noise is reproduced with different parameters, wherein it is determined to reacquire the voice input at least once;
in response to determining not to reacquire the voice input, registering the voice input;
wherein the voice registration device is a server device configured to communicate with a terminal device, and the terminal device is equipped with an audio input device and an audio output device,
wherein the silent environment voice input indicates different registered voice input generated in a silent environment; and
performing voice authentication based on a comparison of different voice input and the registered voice input, where the registered voice input is the voice input acquired during the reproduction of the noise.