Group-Scanning by Face Identification and Real time Emotion detection using Faster R-CNN
K. J. Jayanthi, D. Chitradevi, N Saranya, Sampath Anbukkarasi · 2023
The investigation of face recognition systems carries great significance in the realm of computer vision, as they possess the capability to automatically identify and verify individual faces. These systems are extensively utilized for security purposes in various public and private areas, including banks, malls, hospitals, roads, offices, houses, and government buildings. Several deep learning-based algorithms for facial recognition, real-time emotion detection, and gender classification have been developed in recent years, with most approaches depending on CNN or RCNN. However, these approaches have limitations in processing speed and prediction outputs. This research explores the use of the Faster R-CNN approach for human face detection and realtime emotion recognition to identify individuals and rate their present state. Faster R-CNN has become the leading technology in facial recognition, surpassing other methods in various applications including facial recognition, emotion recognition, and gender classification. Recent advancements in the field of face recognition systems have been achieved by training Faster RCNN models with an accuracy of 93 percent on extensive datasets such as the wider face dataset.