A Novel Deep Learning Based Gender Classification, Expression Recognition and Emoticon Generation

Image is crucial in determining the gender and emotion of an individual in the digital age; however, the issue is evidently with the current methodologies. Implementing deep learning algorithms into image processing would be a more effective strategy. By developing emoticons from the emotions captured in images and snapshots, we hope to bridge the divide in communication. We implemented the Keras framework and evaluated its performance on Tensor Flow utilizing a CNN (Convolutional Neural Network) deep learning algorithm in order to ascertain gender. The objective is to eliminate noise and derive features from the image dataset in order to generate a new dataset that can be utilized to implement CNN. We have utilized the LSTM-RNN (long short-term memory recurrent neural network) to detect emotions and recognize facial expressions. With remarkable accuracy, the algorithm can determine the gender of an individual based on a live webcam feed or an image.