Electronic apparatus and control method thereof

1. An electronic apparatus comprising:

a microphone;

a memory storing a plurality of keyword recognition models; and

a processor, which is coupled with the microphone and the memory, configured to control the electronic apparatus,

wherein the processor is further configured to:

determine whether the electronic apparatus is operating in a first operating state, a second operating state, or a third operating state,

selectively execute a first keyword recognition model, a second keyword recognition model, or a third keyword recognition model among the plurality of keyword recognition models based on operating state information indicating the electronic apparatus is operating in the first operating state, the second operating state, or the third operating state,

based on a first user voice being input through the microphone, identify whether at least one keyword corresponding to the executed first keyword recognition model, the executed second keyword recognition model, or the executed third keyword recognition model is included in the first user voice by using the executed first keyword recognition model, the executed second keyword recognition model, or the executed third keyword recognition model, and

based on at least one keyword identified as being included in the first user voice, perform an operation of the electronic apparatus corresponding to the at least one keyword.

2. The electronic apparatus of claim 1 , wherein the processor is further configured to:

obtain a similarity value between a signal pattern of the at least one keyword and a signal pattern of the first user voice through the executed first keyword recognition model, the executed second keyword recognition model, or the executed third keyword recognition model, and

based on a similarity value of the signal pattern of the at least one keyword exceeding a threshold value among the signal pattern of the first user voice, identify that the first user voice includes the at least one keyword.

3. The electronic apparatus of claim 1 ,

wherein the processor is further configured, based on the electronic apparatus operating in a first operating state in which a conversation system stored in the electronic apparatus is deactivated, to execute a first keyword recognition model corresponding to the first operating state among the plurality of keyword recognition models, and

wherein the processor is further configured, based on the electronic apparatus operating in a second operating state configured to process text corresponding to a second user voice input through the microphone by an activated conversation system, to execute a second keyword recognition model corresponding to the second operating state among the plurality of keyword recognition models.

4. The electronic apparatus of claim 3 , wherein the processor is further configured to:

based on the keyword being not recognized by using the second keyword recognition model executed while operating in the second operating state, input the first user voice to an automatic speech recognition (ASR) module of the conversation system to obtain text corresponding to the first user voice, and

provide a response to the first user voice based on the text corresponding to the first user voice.

5. The electronic apparatus of claim 3 , wherein the processor is further configured to, based on the electronic apparatus operating in a third operating state configured to recognize a second user voice input through the microphone by the activated conversation system, through an automatic speech recognition (ASR) module of the conversation system, stop executing the plurality of keyword recognition models and obtain text corresponding to the second user voice through the ASR module.

6. The electronic apparatus of claim 3 , wherein the processor is further configured to, based on a keyword including a meaning of requesting a predefined operation being identified as being included in the first user voice by using the first keyword recognition model executed while operating in the first operating state, control to perform the predefined operation.

7. The electronic apparatus of claim 3 , wherein the processor is further configured to:

based on a keyword including a meaning of requesting a next or previous operation being identified as being included in the first user voice by using the first keyword recognition model executed while operating in the first operating state, identify an operation performed by the electronic apparatus by the conversation system before the first keyword recognition model identifies that the keyword is included in the first user voice, on a basis of conversation history information, and

control to perform an operation corresponding to the next or previous operation of the operation performed by an identified electronic apparatus.

8. The electronic apparatus of claim 3 , wherein the processor is further configured to, based on a keyword including a meaning of stopping an operation being identified as being included in the first user voice by using the second keyword recognition model executed while operating in the second operating state, control to stop processing with respect to text corresponding to the second user voice by the conversation system.

9. The electronic apparatus of claim 3 , wherein the processor is further configured to, based on a keyword including a meaning of repeating an operation being identified as being included in the first user voice by using the second keyword recognition model executed while operating in the second operating state, control to stop processing the text corresponding to the second user voice performed by the conversation system and to repeat processing of the text based on conversation history information.

10. A method of controlling an electronic apparatus comprising:

determining whether the electronic apparatus is operating in a first operating state, a second operating state, or a third operating state;

selectively executing a first recognition model, a second keyword recognition model, or a third keyword recognition model among a plurality of keyword recognition models based on operating state information indicating the electronic apparatus is operating in the first operating state, the second operating state, or the third operating state;

based on a first user voice being input through a microphone, identifying whether at least one keyword corresponding to the executed first keyword recognition model, the executed second keyword recognition model, or the executed third keyword recognition model is included in the first user voice by using the executed first keyword recognition model, the executed second keyword recognition model, or the executed third keyword recognition model; and

based on at least one keyword identified as being included in the first user voice, performing an operation of the electronic apparatus corresponding to the at least one keyword.

11. The method of claim 10 , wherein the identifying comprises:

obtaining a similarity value between a signal pattern of the at least one keyword and a signal pattern of the first user voice through the executed first keyword recognition model, the executed second keyword recognition model, or the executed third keyword recognition model; and

based on a similarity value of the signal pattern of the at least one keyword exceeding a threshold value among the signal pattern of the first user voice, identifying that the first user voice includes the at least one keyword.

12. The method of claim 10 , wherein the executing comprises:

based on the electronic apparatus operating in a first operating state in which a conversation system stored in the electronic apparatus is deactivated, executing a first keyword recognition model corresponding to the first operating state among the plurality of keyword recognition models, and

based on the electronic apparatus operating in a second operating state that processes text corresponding to a second user voice input through the microphone by an activated conversation system, executing a second keyword recognition model corresponding to the second operating state among the plurality of keyword recognition models.

13. The method of claim 12 , wherein the recognizing comprises:

based on the keyword being not recognized by using the second keyword recognition model executed while operating in the second operating state, inputting the first user voice to an automatic speech recognition (ASR) module of the conversation system to obtain text corresponding to the first user voice, and

providing a response to the first user voice based on the text corresponding to the first user voice.

14. The method of claim 12 , wherein the executing comprises, based on the electronic apparatus operating in a third operating state configured to recognize a second user voice input through the microphone by the activated conversation system, through an automatic speech recognition (ASR) module of the conversation system, stopping execution of the plurality of keyword recognition models and obtaining text corresponding to the second user voice through the ASR module.

15. The method of claim 12 , wherein the performing comprises, based on a keyword including a meaning of requesting a predefined operation being identified as being included in the first user voice by using the first keyword recognition model executed while operating in the first operating state, controlling to perform the predefined operation.

16. The method of claim 12 , wherein the performing comprises:

based on a keyword including a meaning of requesting a next or previous operation being identified as being included in the first user voice by using the first keyword recognition model executed while operating in the first operating state, identifying an operation performed by the electronic apparatus by the conversation system before the first keyword recognition model identifies that the keyword is included in the first user voice, on a basis of conversation history information; and

performing an operation corresponding to the next or previous operation of the operation performed by an identified electronic apparatus.

17. The method of claim 12 , wherein the performing comprises, based on a keyword including a meaning of stopping an operation being identified as being included in the first user voice by using the second keyword recognition model executed while operating in the second operating state, stopping processing with respect to text corresponding to the second user voice by the conversation system.

18. The method of claim 12 , wherein the performing comprises, based on a keyword including a meaning of repeating an operation being identified as being included in the first user voice by using the second keyword recognition model executed while operating in the second operating state, stopping processing the text corresponding to the second user voice performed by the conversation system and to repeat processing of the text based on conversation history information.