Analysis of Facial Expression using Deep Learning Techniques

• Published in: 2023 IEEE International Conference on Integrated Circuits and Communication Systems (ICICACS) pp. 1 - 8
• Main Authors: C Patil, Priyadarshini, R K, Ashwin, Kumar G, Arvind, Bhaskar, M, N, Rajesh
• Format: Conference Proceeding
• Language: English
• Published: IEEE 24.02.2023
• Subjects: accurate; Brightness; deep learning; Emotion recognition; face; Face recognition; genetic facial regression; Genetics; Image color analysis; Mouth; Shape
• DOI: 10.1109/ICICACS57338.2023.10100110

Research into facial recognition has been one of the most intriguing and extensive areas of study for decades. Because of its importance in reading and conveying other people's emotions, the face often becomes the focal point of conversations. A facial recognition system's ability to identify an individual from a digital image or video is useful for a wide variety of security-related purposes, including surveillance, general identity verification, criminal-justice Systems, image database investigations, Smart Card applications, multi-media settings with adaptive human-computer interfaces, video indexing, gender classification, facial feature recognition, and tracking. While being less reliable than fingerprint and iris identification, face recognition has seen widespread adoption because of its contactless, non-invasive nature. Light, emotion, and posture all have a role in making it harder to recognise a person's face. A computer model of face recognition is difficult to build because of the problem space's complexity and possible multidimensionality. Extensive attempts have been made to build accurate and reliable face recognition systems. The high dimensionality of a facial image is necessary for detecting minute variations in features, but this also means that the calculations required to classify the image take a long time. It's possible that lowering the image resolution will reduce the time needed for recognition processing. Facial recognition systems have been studied for quite some time, but there is always opportunity for improvement. These results show that current face recognition algorithms have matured to a considerable extent while operating in a constrained context. When put to the test in the real world, however, these technologies do not show a significant performance boost in all of the frequent cases faced by applications. Genetic face recognition is necessary when only the most cutting-edge method of facial identification will do. There are many potential applications for genetic face recognition technology; home and corporate security are only two examples. When two people have very similar faces, it's easy to tell whether they're related. It's also called "face-of-relatives" verification. Those that have a common ancestor are said to share a genetic makeup, whereas those who come from distinct families are said to lack such a makeup.