Cross-language communication voice robot interaction method and cross-language communication voice robot interaction device

Abstract

The invention discloses a cross-language communication voice robot interaction method and a cross-language communication voice robot interaction device. The method comprises the steps of acquiring first voice information of an external interaction object; inputting the first voice information into a plurality of single language recognizers connected in parallel, and outputting a plurality of voicerecognition results; comparing each voice recognition result with a corpus of multiple languages respectively, calculating a plurality of likelihood probabilities of the multiple languages corresponding to each voice recognition result, and selecting a voice recognition result with the maximum likelihood probability value of the maximum likelihood probability from the voice recognition results asa voice recognition result of the first voice information according to the maximum likelihood probability corresponding to each voice recognition result; and outputting second voice information responding the first voice information according to the voice recognition result of the first voice information by adopting the language of the voice recognition result with the maximum probability value of the maximum likelihood probability. According to the method, a voice robot can realize the technical effects of quick response and answer under a cross-language communication scene.

Description

technical field [0001] The invention relates to the field of voice recognition, in particular to a voice robot interaction method and device for cross-language communication. Background technique [0002] This section is intended to provide a background or context to embodiments of the invention that are recited in the claims. The descriptions herein are not admitted to be prior art by inclusion in this section. [0003] With the rapid development of artificial intelligence and speech recognition technology, more and more voice robots appear in people's daily life, such as shopping guide robots in shopping malls and lobby robots in banks. The response speed of these voice robots to external voice communication directly affects the user's interactive experience. [0004] At present, existing voice robots use a single-language recognizer for voice communication, which can only realize voice communication in a single language, and cannot cope with voice communication in diffe...

AI Technical Summary

Problems solved by technology

[0008] An embodiment of the present invention provides a voice robot interaction method for cross-language communication, which is used to solve the technical problem that the existing voice robot uses a single-language recognizer for voice communication and is prone to errors when facing a cross-language communication scene. The method includes: Acquiring the first voice information of the external interactive object; inputting the first voice information into multiple monolingual recognizers connected in parallel, and outputting multiple voice recognition results, wherein each single language recognizer is used to recognize voice information of one language ; Comparing each speech recognition result with the corpus of multiple languages ​​respectively, calculating multiple likelihood probabilities corresponding to each speech recognition result in multiple languages, and obtaining the maximum likelihood probability corresponding to each speech recognition result; according to each The maximum likelihood probability corresponding to each speech recognition result, from a plurality of speech recognition results, select the speech recognition result with the largest probability value of the maximum likelihood probability as the speech recognition result of the first speech information; according to the speech of the first speech information As a result of the recognition, use the language of the speech recognition result with the largest probability value of the maximum likelihood probability, and output the second speech information in response to the first speech information

[0009] The embodiment of the present invention also provides a voice robot interaction device for cross-language communication, which is used to solve the technical problem that the existing voice robot uses a single-language recognizer for voice communication and is prone to make mistakes when facing a cross-language communication scene. The device includes : a voice collection unit, used to acquire the first voice information of the external interactive object; a voice recognition unit, used to input the first voice information into a plurality of single-language recognizers connected in parallel, and output a plurality of voice recognition results, wherein each A single-language recognizer is used to recognize speech information in one language; a speech recognition result probability calculation unit is used to compare each speech recognition result with a corpus of multiple languages, and calculate each speech recognition result corresponding to multiple languages. A plurality of likelihood probabilities, and obtain the maximum likelihood probability corresponding to each speech recognition result; the speech recognition result selection unit is used to select from a plurality of speech recognition results according to the maximum likelihood probability corresponding to each speech recognition result The speech recognition result with the maximum probability value of the maximum likelihood probability is used as the speech recognition result of the first speech information; the speech output unit is used to adopt the speech with the maximum probability value of the maximum likelihood probability according to the speech recognition result of the first speech information The language of the recognition result, and output the second voice information in response to the first voice information

[0010] The embodiment of the present invention also provides a computer device, which is used to solve the technical problem that the existing voice robot uses a single language recognizer for voice communication, and is prone to make mistakes when facing a cross-language communication scene. The computer device includes a memory, a processor and A computer program stored on a memory and operable on a processor, when the processor executes the computer program, the voice robot interaction method for the above-mentioned cross-language communication is realized

Images