Speech emotion recognition method and apparatus

A method, comprising:

• determining, based on a first neural network model, a plurality of pieces of emotional state information corresponding to a plurality of speech frames comprised in a current utterance in a target dialog, wherein the plurality of speech frames have a one-to-one correspondence with the plurality of pieces of emotional state information, and each piece of emotional state information represents an emotional state corresponding to a corresponding speech frame;
• performing a statistical operation on the plurality of pieces of emotional state information to obtain a statistical result, wherein the statistical result corresponds to the current utterance;
• and
• determining emotional state information corresponding to the current utterance, the determining being based on a second neural network model, the statistical result corresponding to the current utterance, and n-1 statistical results corresponding to n-1 utterances before the current utterance;
• and
• wherein the n-1 utterances have a one-to-one correspondence with the n-1 statistical results, a statistical result corresponding to each utterance in the n-1 utterances is obtained by performing a statistical operation on a plurality of pieces of emotional state information corresponding to a plurality of speech frames comprised in the respective utterance, the n-1 utterances belong to the target dialog, and n is an integer greater than 1.

An apparatus, comprising:

• a non-transitory memory, configured to store programs;
• and
• a processor, configured to execute the programs stored in the memory, wherein when at least one program stored in the memory is executed, the processor is configured to:
• determine, based on a first neural network model, a plurality of pieces of emotional state information corresponding to a plurality of speech frames comprised in a current utterance in a target dialog, wherein the plurality of speech frames have a one-to-one correspondence with the plurality of pieces of emotional state information, and each piece of emotional state information represents an emotional state corresponding to the corresponding speech frame;
• perform a statistical operation on the plurality of pieces of emotional state information to obtain a statistical result, wherein the statistical result corresponds to the current utterance;
• and
• determine emotional state information corresponding to the current utterance, the determining being based on a second neural network model, the statistical result corresponding to the current utterance, and n-1 statistical results corresponding to n-1 utterances before the current utterance;
• and
• wherein the n-1 utterances have a one-to-one correspondence with the n-1 statistical results, a statistical result corresponding to each utterance in the n-1 utterances is obtained by performing statistical operation on a plurality of pieces of emotional state information corresponding to a plurality of speech frames comprised in the respective utterance, the n-1 utterances belong to the target dialog, and n is an integer greater than 1.

A non-transitory computer readable storage medium storing executable instructions, the executable instructions including instructions for:

• determining, based on a first neural network model, a plurality of pieces of emotional state information corresponding to a plurality of speech frames comprised in a current utterance in a target dialog, wherein the plurality of speech frames have a one-to-one correspondence with the plurality of pieces of emotional state information, and each piece of emotional state information represents an emotional state corresponding to a corresponding speech frame;
• performing a statistical operation on the plurality of pieces of emotional state information to obtain a statistical result, wherein the statistical result corresponds to the current utterance;
• and
• determining emotional state information corresponding to the current utterance, the determining being based on a second neural network model, the statistical result corresponding to the current utterance, and n-1 statistical results corresponding to n-1 utterances before the current utterance;
• and
• wherein the n-1 utterances have a one-to-one correspondence with the n-1 statistical results, a statistical result corresponding to each utterance in the n-1 utterances is obtained by performing a statistical operation on a plurality of pieces of emotional state information corresponding to a plurality of speech frames comprised in the respective utterance, the n-1 utterances belong to the target dialog, and n is an integer greater than 1.