Emotion Detection Using Deep Learning Algorithm

Automatic emotion detection is a prime task in computerized human behaviour analysis. The proposed system an automatic using convolution neural network. end-to-end CNN therefore named as ENet. Keeping mind the computational efficiency, deep network makes use of trained weight parameters MobileNet to initialize On top last layer ENet, authors place global average pooling make it independent input image size. ENet validated for two benchmark datasets: Cohn-Kanade+ (CK+) and Japanese female facial expression (JAFFE). experimental results show that outperforms other existing methods detection.