Development of an interest detection system based on facial analysis using deep learning

In the marketing research, one of the most exciting, innovative, and promising trends is quantification of customer interest. The customer satisfaction survey, which is a traditional approach to quantify customer interest, has come to be considered as an invasive method in recent years. Recording customer interest by a salesperson who observes customers' behavior during the advertisement watching or shopping phase is another approach. However, this task requires specific skills for every salesperson, and each observer may interpret customer behaviors differently. Consequently, there is a critical need to develop non-invasive, objective, and quantitative tools for monitoring customer interest. This study presents a deep learning-based system for monitoring customer behavior specifically for detection of interest. The proposed system first measures customer attention through head pose estimation. For those customers whose heads are oriented toward the advertisement or the product of interest, the system further analyzes the facial expressions and reports customers' interest. The proposed system starts by detecting frontal face poses; facial components important for facial expression recognition are then segmented and an iconized face image is generated; finally, facial expressions are analyzed using the confidence values of obtained iconized face image combined with the raw facial images. This approach fuses local part-based features with holistic facial information for robust facial expression recognition. The system is also tracked human faces along the video frame by labeling the faces. The facial expressions of each customer are stored for a certain period of time; at the end of this period, the result of whether the customer is related to the product or advirtesement is notified. With the proposed processing pipeline, using a basic imaging device, such as a webcam, head pose estimation and facial expression recognition is possible. The proposed pipeline can be used to monitor emotional response of focus groups to various ideas, pictures, sounds, words, and other stimuli.