Semantic Feature Extraction and Deep Convolutional Neural Network-based Face Sentimental Analysis

Abstract

Police and government agencies make use of facial recognition technology in order to determine the truth about the criminal. Though this might seem skeptical, it requires the support of the public in order to come into action. In this regard the media plays a major role in molding the public agenda and attracting sentiments and attitudes towards this topic. In this work, various perspectives are taken into consideration in order to determine the impact of social media on the public, police and government with the help of face recognition technology. A total of 443 videos have been analyzed and the outcome showed to be positive for this technology to be incorporated. Close examination of emotional language indicated several levels of anticipation and surprise along with fear and sadness. It is worth noting that trust is in emotion expressed in low levels only. Deep learning based CNN technique is used for categorization. Based on the information obtained and by incorporating new methodologies, conclusions are drawn, strategies are incorporated and recorded.