Abstract:

A deep learning approach was proposed in this study for estimating age, emotion expression, and gender from a real-time video source without using facial landmarks or other geometric calculations of café features. For image classification, a convolutional neural networks (CNNs) pre-trained and used on ImageNet from Caffe (Convolutional Architecture for Quick Feature Embedding), a modifiable platform for state-of-the-art deep learning algorithms and a set of reference models. The (you only look once v3) YOLOv3 algorithm was employed for such purposes having a desirable abilities to serve the required purpose. Deep Convolutional Neural Network (DCNN) features are used to propose a framework for automatically understanding facial expressions. The suggested model focuses on understanding an individual's facial expressions from a single image.

Keywords: