Bimodal astronaut emotion recognition method based on facial expressions and voices in video

The invention discloses a bimodal astronaut emotion recognition method based on facial expressions and voices in a video, and the method comprises the steps: intercepting a video frame from RAVDESS for a facial expression mode in the video, judging whether the video frame contains a human face or not, and cutting an original image containing the human face to a facial region; editing the original face image into a face image in a simulated space environment; inputting the preprocessed facial image into a convolutional neural network ATTCNN containing an attention mechanism to obtain an astronaut facial expression modal emotion recognition result; for a voice mode in the video, audio data in the RAVDESS are superposed with random noise to obtain a voice signal in a simulated space environment; inputting the signal into a voice enhancement network to obtain a clean voice signal, extracting MFCC features of the clean voice signal, and inputting the MFCC features into a CNNLSTM network to obtain an astronaut voic