The Intel Realsense Depth-Camera Performance for Real-Time Customer Satisfaction Analysis using Facial Expression Detection

Autor: James Purnama, Jason Yapri, Stefanus Oliver, Maulahikmah Galinium, Tommy Winarta
Rok vydání: 2019
Předmět:
Zdroj: Journal of Physics: Conference Series. 1175:012076
ISSN: 1742-6596
1742-6588
DOI: 10.1088/1742-6596/1175/1/012076
Popis: Customer satisfaction evaluation is an important process in business development. Unfortunately, doing it manually by using survey or by phone will take up the time and resources that can actually be reallocated. In this research, we are going to make an automated customer satisfaction analysis tool by using Intel RealSense. The depth sensor and its emotion recognition API is enable real-time 3-D customer emotion recognition monitoring accurately. Emotions such as anger, disgust, fear, joy, sadness, and surprise that is expressed by the customer is logged. Then, the captured emotions are processed and analyzed by the computer. Finally, the summarized survey presents an interactive data in form of charts that can show the customer service's performance. As a result, the data that is produced can be used to improve the performance of the customer service representative. In this research, experiments show that while Intel RealSense can capture customer's emotions well most of the time, there are still some cases in which it couldn't present the accurate result due to the limitation in the API. Therefore, further testing need to be done that also takes several factors e.g. head accessories, glasses into account to increase the credibility of the output from the program.
Databáze: OpenAIRE