Speech noise reduction model training method and apparatus, speech scoring method and apparatus, device, storage medium, and program product

An electronic device, comprising:

• a memory, configured to store a computer executable instruction;
• and
• a processor, configured to execute the computer executable instruction stored in the memory and cause the electronic device to perform a speech scoring method including:
• receiving speech information and associated reference speech text via a speech input interface, wherein the reference speech text has a corresponding reference pronunciation and the speech information represents an audio signal of a person reading the associated reference speech text to the speech input interface;
• performing noise reduction processing on the speech information based on a speech noise reduction model to obtain noise-reduced speech information;
• performing speech recognition on the noise-reduced speech information to recognize text in the noise-reduced speech information and acoustic features associated with the speech information;
• predicting a pronunciation score for indicating pronunciation similarity between the speech information and the reference pronunciation corresponding to the reference speech text based on (i) content differences between the recognized text and the associated reference speech text and (ii) pronunciation differences between the acoustic features of the audio signal and the reference pronunciation;
• and
• presenting the pronunciation score on the speech input interface in response to a trigger operation on a speech end function item in the speech input interface by the person.

A non-transitory computer-readable storage medium, storing a computer executable instruction that, when executed by a processor of an electronic device, causes the electronic device to perform a speech scoring method including:

• receiving speech information and associated reference speech text via a speech input interface, wherein the reference speech text has a corresponding reference pronunciation and the speech information represents an audio signal of a person reading the associated reference speech text to the speech input interface;
• performing noise reduction processing on the speech information based on a speech noise reduction model to obtain noise-reduced speech information;
• performing speech recognition on the noise-reduced speech information to recognize text in the noise-reduced speech information and acoustic features associated with the speech information;
• predicting a pronunciation score for indicating pronunciation similarity between the speech information and the reference pronunciation corresponding to the reference speech text based on (i) content differences between the recognized text and the associated reference speech text and (ii) pronunciation differences between the acoustic features of the audio signal and the reference pronunciation;
• and
• presenting the pronunciation score on the speech input interface in response to a trigger operation on a speech end function item in the speech input interface by the person.