Method and apparatus for outputting information

1. A method for outputting information, the method comprising:

acquiring voice information received within a preset time period before a device is awakened, wherein the device is provided with a wake-up model for outputting preset response information when a preset wake-up word is received;

performing speech recognition on the voice information to obtain a recognition result;

extracting feature information of the voice information in response to determining that the recognition result does not include the preset wake-up word;

generating a counterexample training sample based on the feature information; and

training the wake-up model using the counterexample training sample, and outputting a trained wake-up model, wherein the training is performed by using a new wake-up word included in the voice information in the counterexample training sample as an input, and using null information as an expected output.

2. The method according to claim 1 , wherein the training the wake-up model using the counterexample training sample and outputting the trained wake-up model comprises:

inputting the feature information of the voice information comprising the preset wake-up word into the trained wake-up model, and determining whether the preset response information is output; and

outputting the trained wake-up model in response to outputting the preset response information.

3. The method according to claim 2 , wherein the method further comprises:

in response to not outputting the preset response information, outputting alarm information.

4. The method according to claim 1 , wherein the training the wake-up model using the counterexample training sample, and outputting the trained wake-up model comprises:

determining a number of the counterexample training samples; and

training the wake-up model with the counterexample training samples in response to the number of the counterexample training samples being greater than or equal to a preset number threshold.

5. The method according to claim 4 , wherein the acquiring voice information received within a predetermined time period before a device is awakened comprises:

in response to the device being awakened by the preset wake-up word, acquiring the voice information received within the preset time period before the device is awakened.

6. A server comprising:

one or more processors;

a storage apparatus storing one or more programs,

wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to perform operations, the operations comprising:

acquiring voice information received within a preset time period before a device is awakened, wherein the device is provided with a wake-up model for outputting preset response information when a preset wake-up word is received;

performing speech recognition on the voice information to obtain a recognition result;

extracting feature information of the voice information in response to determining that the recognition result does not include the preset wake-up word;

generating a counterexample training sample based on the feature information; and

training the wake-up model using the counterexample training sample, and outputting a trained wake-up model, wherein the training is performed by using a new wake-up word included in the voice information in the counterexample training sample as an input, and using null information as an expected output.

7. The server according to claim 6 , wherein the training the wake-up model using the counterexample training sample, and outputting the trained wake-up model comprises:

inputting the feature information of the voice information comprising the preset wake-up word into the trained wake-up model, and determining whether the preset response information is output; and

outputting the trained wake-up model in response to outputting the preset response information.

8. The server according to claim 7 , wherein the operations further comprises:

in response to not outputting the preset response information, outputting alarm information.

9. The server according to claim 6 , wherein the training the wake-up model using the counterexample training sample, and outputting the trained wake-up model comprises:

determining a number of the counterexample training samples; and

training the wake-up model with the counterexample training samples in response to the number of the counterexample training samples being greater than or equal to a preset number threshold.

10. The server according to claim 9 , wherein the acquiring voice information received within a predetermined time period before a device is awakened comprises:

in response to the device being awakened by the preset wake-up word, acquiring the voice information received within the preset time period before the device is awakened.

11. A non-transitory computer readable medium storing a computer program, wherein the computer program, when executed by a processor, causes the processor to perform operations, the operations comprising:

acquiring voice information received within a preset time period before a device is awakened, wherein the device is provided with a wake-up model for outputting preset response information when a preset wake-up word is received;

performing speech recognition on the voice information to obtain a recognition result;

extracting feature information of the voice information in response to determining that the recognition result does not include the preset wake-up word;

generating a counterexample training sample based on the feature information; and

training the wake-up model using the counterexample training sample, and outputting a trained wake-up model, wherein the training is performed by using a new wake-up word included in the voice information in the counterexample training sample as an input, and using null information as an expected output.

12. The non-transitory computer readable medium according to claim 11 , wherein the training the wake-up model using the counterexample training sample and outputting the trained wake-up model comprises:

inputting the feature information of the voice information comprising the preset wake-up word into the trained wake-up model, and determining whether the preset response information is output; and

outputting the trained wake-up model in response to outputting the preset response information.

13. The non-transitory computer readable medium according to claim 12 , wherein the operations further comprises:

in response to not outputting the preset response information, outputting alarm information.

14. The non-transitory computer readable medium according to claim 11 , wherein the training the wake-up model using the counterexample training sample, and outputting the trained wake-up model comprises:

determining a number of the counterexample training samples; and

training the wake-up model with the counterexample training samples in response to the number of the counterexample training samples being greater than or equal to a preset number threshold.

15. The non-transitory computer readable medium according to claim 14 , wherein the acquiring voice information received within a predetermined time period before a device is awakened comprises:

in response to the device being awakened by the preset wake-up word, acquiring the voice information received within the preset time period before the device is awakened.