Method and apparatus for translating speech

1. A method for translating speech, comprising:

recognizing received to-be-recognized speech of a source language to obtain a recognized text;

concatenating the obtained recognized text after a to-be-translated text, to form a concatenated to-be-translated text;

inputting the concatenated to-be-translated text into a pre-trained discriminant model to obtain a discrimination result for characterizing whether the concatenated to-be-translated text is to be translated, wherein the discriminant model is used to characterize a corresponding relationship between a text and a discrimination result corresponding to the text, and the discrimination result corresponding to the text is a positive discrimination result for indicating the text being to be translated separately or a negative discrimination result for indicating the text being not to be translated separately;

in response to the positive discrimination result being obtained, translating the concatenated to-be-translated text to obtain a translation result of a target language, and outputting the translation result.

2. The method according to claim 1 , wherein before outputting the translation result, the method further comprises:

clearing the concatenated to-be-translated text.

3. The method according to claim 1 , further comprising:

in response to the negative discrimination result being obtained, using the concatenated to-be-translated text as the to-be-translated text and performing recognition on a next to-be-recognized speech in real time.

4. The method according to claim 1 , wherein the discriminant model is pre-trained through:

determining a network structure of an initial neural network and initializing network parameters of the initial neural network;

acquiring a training sample set, wherein a training sample comprises a sample text and corresponding annotation information, the annotation information is used to characterize whether the sample text is fit for being translated separately;

training the initial neural network using a machine learning method, by taking a sample text and annotation information in the training sample of the training sample set as an input and a desired output of the initial neural network respectively; and

determining the trained initial neural network as a pre-trained discriminant model.

5. The method according to claim 4 , wherein the initial neural network comprises an input layer, a hidden layer, a fully connected layer, and a classifier layer.

6. An apparatus for translating speech, comprising:

at least one processor; and

a memory storing instructions, the instructions when executed by the at least one processor, cause the at least one processor to perform operations, the operations comprising:

recognizing received to-be-recognized speech of a source language to obtain a recognized text;

concatenating the obtained recognized text after a to-be-translated text, to form a concatenated to-be-translated text;

inputting the concatenated to-be-translated text into a pre-trained discriminant model to obtain a discrimination result for characterizing whether the concatenated to-be-translated text is to be translated, wherein the discriminant model is used to characterize a corresponding relationship between a text and a discrimination result corresponding to the text, and the discrimination result corresponding to the text is a positive discrimination result for indicating the text being to be translated separately or a negative discrimination result for indicating the text being not to be translated separately; and

translating, in response to the positive discrimination result being obtained, the concatenated to-be-translated text to obtain a translation result of a target language, and outputting the translation result.

7. The apparatus according to claim 6 , wherein the operations further comprise:

clearing the concatenated to-be-translated text before outputting the translation result.

8. The apparatus according to claim 6 , wherein the operations further comprise:

in response to the negative discrimination result being obtained, using the concatenated to-be-translated text as the to-be-translated text and performing recognition on a next to-be-recognized speech in real time.

9. The apparatus according to claim 6 , wherein the discriminant model is pre-trained through:

determining a network structure of an initial neural network and initializing network parameters of the initial neural network;

acquiring a training sample set, wherein a training sample comprises a sample text and corresponding annotation information, the annotation information is used to characterize whether the sample text is fit for being translated separately;

training the initial neural network using a machine learning method, by taking a sample text and annotation information in the training sample of the training sample set as an input and a desired output of the initial neural network respectively; and

determining the trained initial neural network as a pre-trained discriminant model.

10. The apparatus according to claim 9 , wherein the initial neural network comprises an input layer, a hidden layer, a fully connected layer, and a classifier layer.

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

recognizing received to-be-recognized speech of a source language to obtain a recognized text;

concatenating the obtained recognized text after a to-be-translated text, to form a concatenated to-be-translated text;

inputting the concatenated to-be-translated text into a pre-trained discriminant model to obtain a discrimination result for characterizing whether the concatenated to-be-translated text is to be translated, wherein the discriminant model is used to characterize a corresponding relationship between a text and a discrimination result corresponding to the text, and the discrimination result corresponding to the text is a positive discrimination result for indicating the text being to be translated separately or a negative discrimination result for indicating the text being not to be translated separately;

in response to the positive discrimination result being obtained, translating the concatenated to-be-translated text to obtain a translation result of a target language, and outputting the translation result.

12. The medium according to claim 11 , wherein before outputting the translation result, the operations further comprise:

clearing the concatenated to-be-translated text.

13. The medium according to claim 11 , wherein the operations further comprise:

in response to the negative discrimination result being obtained, using the concatenated to-be-translated text as the to-be-translated text and performing recognition on a next to-be-recognized speech in real time.

14. The medium according to claim 11 , wherein the discriminant model is pre-trained through:

determining a network structure of an initial neural network and initializing network parameters of the initial neural network;

acquiring a training sample set, wherein a training sample comprises a sample text and corresponding annotation information, the annotation information is used to characterize whether the sample text is fit for being translated separately;

training the initial neural network using a machine learning method, by taking a sample text and annotation information in the training sample of the training sample set as an input and a desired output of the initial neural network respectively; and

determining the trained initial neural network as a pre-trained discriminant model.

15. The Medium according to claim 14 , wherein the initial neural network comprises an input layer, a hidden layer, a fully connected layer, and a classifier layer.

16. The method according to claim 2 , wherein after clearing the concatenated to-be-translated text, setting the to-be-translated text to be empty and performing recognition on a next to-be-recognized speech in real time.