Design of an Algorithm to Recognize and Analyze Personal Behavior Cues

Abstract

Affective computing is an active area of research which is incorporation of multidisciplinary newlinefields such as artificial intelligence, cognitive science, natural language processing and newlinesocial sciences. Emotion recognition, human behavior recognition and sentiment analysis newlineresearch areas fall under the umbrella of affective computing. With technological advancement and increased smartphone and social media usage, a newlinetremendous amount of big data is getting generated. Along with text data, video and audio newlinedata is also available to scientists for research purposes. The use of video data for behavior newlinerecognition provides multiple modalities like facial expressions, speech modulations and newlinecontext-aware text information for processing. Most of the state-of-the-art approaches for newlinebehavior recognition focus on single modalities or bimodalities. These frameworks newlinedemonstrate a limitation in terms of overall accuracy, performance, and robustness. Detecting and recognizing personalized behavior cues is one of the challenging tasks. Affect newlinesensing is one of the first steps in identifying the behavior of the person. Affect is a process newlineof recognizing human emotions with the help of signals, data, and patterns.