Automated Analysis and Behavioural Prediction of Interview Performance using Computer Vision

The behavioral interview is an interviewing technique that employers use to evaluate a potential employee. To accelerate the recruiting process and make recruiting on a simple scale while improving the applicant experience. This proposed model automatically analyses and predicts the non-verbal characteristics of an interviewee and also provides relevant feedback. This video-based analysis includes facial expressions and head gestures. To implement such a model, various regression, classification, and deep learning algorithms are generally used. By using the Random Forest algorithm and deep-learning frameworks we are predicting the consequences and prescribing feedback based on the behavior of the candidates. Furthermore, we will test the hypothesis "The first impression is the best impression," as well as output the attributes that the candidate lacked throughout the interview and provide essential suggestions on how to enhance performance.