A deep learning approach for facial emotions recognition using principal component analysis and neural network techniques

In this work, advanced facial emotions are recognized using Neural network-based (NN) PCA methodology. The earlier models are cannot detect facial emotions with moving conditions but the CCTV and other advanced applications are mostly depending on moving object-based emotion recognition. The blurring, mask, and moving object-based facial image are applied to the training process, and at the testing condition, real-time facial images are applied. The PCA is extracting features and pre-processing the images with NN deep learning process. The proposed facial emotion recognition model is most useful for advanced applications. The design is finally verified through confusion matrix computations and gets measures like accuracy 98.34%, sensitivity 98.34% Recall 97.34%, and F score 98.45%. These output results compete with present models and outperformance the methodology.