Face Recognition (Engineering)

Some individuals with aphantasia, an absent or reduced ability to form visual imagery, report face recognition difficulties, and studies have shown aphantasics exhibit impaired face recognition performance on the Cambridge Face Memory Test. While this impaired performance may be due to poor face memory, it may instead, or in addition, be due to difficulties in other face processing domains (such as face perception or face matching). The Cambridge Face Memory Test, Oxford Face Matching Test and a novel analysis strategy were used to obtain independent measures of face memory, face matching and face perception from 34 individuals with aphantasia and an age-matched control group. Results showed that individuals with aphantasia exhibit impairments in face memory and face perception, but not face matching. This study therefore provides information about the specific face processing mechanisms that may be impaired in aphantasia.

The development of artificial intelligence in facial emotion recognition (FER) is rapidly growing and has been widely applied in various fields. Deep learning (DL) techniques with evolutionary algorithms have become the preferred choice for solving various security, health, gaming, and other related problems. This research proposes the use of a genetic algorithm (GA) as the main method to optimize hyperparameters in the convolutional neural network (CNN) model for FER. The required computation time is approximately 37 hours 57 minutes 55 seconds, with generation 3 taking the longest time at around 16 hours 45 minutes 4 seconds. However, generation 3 achieved an accuracy of 76.11%, which is the highest compared to other generations. The results indicate that the more generations are involved, the higher the achievable accuracy. Furthermore, the proposed CNN-GA model in this study outperforms previous models that have been examined. Thus, this study makes a significant contribution to improving the understanding of using GAs to optimize the performance of CNN models for FER.

Facial Authentication System is a biometric technology that identifies or verifies a person using facial features. This research focuses on designing a secure and efficient authentication system using computer vision and machine learning techniques. The system captures facial images, detects faces, extracts unique features, and compares them with stored data for authentication. The proposed system aims to provide high accuracy, fast processing, and enhanced security compared to traditional methods like passwords or ID cards. The Face Recognition Attendance System represents an innovative solution at the intersection of artificial intelligence and attendance management. Leveraging Python open-source libraries such as OpenCV and NumPy, alongside machine learning techniques, the system aims to streamline the attendance tracking process. By employing face detection algorithms and feature extraction methods, the system identifies individuals, records their attendance, and updates the database in real-time. The system architecture incorporates components like a front-end web application, real-time prediction module, registration form module, reporting module, Redis database, Streamlit framework, Insightface library, and AWS deployment. Through continuous improvement and innovation, the system enhances user experience, accuracy, and efficiency in attendance management. Facial recognition technology has emerged as a vital component in modern security systems. This paper presents the design and implementation of a low-cost, real-time facial recognition and tracking system using an Arduino microcontroller, a USB camera, and Python-based image processing. The system captures live video, detects faces using the OpenCV library, and compares them against a pre-stored database. Upon successful recognition, the Arduino triggers an output device such as a servo-controlled door lock or an alarm. The proposed system demonstrates that affordable hardware can be integrated with open-source software to achieve reliable facial authentication for small-scale security applications.

Face recognition systems play a crucial role in security, surveillance, and authentication applications. However, traditional deep learning-based models, particularly Convolutional Neural Networks (CNNs), often struggle with issues such as varying lighting conditions, occlusions, and high computational costs. This paper proposes an Adaptive Resonance Theory (ART)-based face recognition framework that enhances recognition robustness and computational efficiency. Unlike CNNs, ART enables incremental learning without requiring retraining, making it suitable for realtime applications. The study evaluated the proposed system on three benchmark datasets: LFW, Yale, and ORL. Experimental results indicate that the ART-based model achieved an average accuracy of 96.2%, outperforming CNN-based models (93.5%) while reducing recognition time by 25%. Additionally, ART demonstrated superior adaptability, maintaining recognition accuracy above 94% even under occlusion and low-light conditions. These findings confirm the effectiveness of ART-based face recognition for security, access control, and innovative surveillance applications. Future research will focus on integrating ART with deep learning techniques for enhanced performance in large-scale datasets.

Prioritizing advanced security, the infusion of AI Mediapipe into the realm of smart door access not only simplifies access control but also guarantees unmatched protection, reshaping the expectation for safety and reliability. Implementing AI Mediapipe facilitates real-time gesture recognition, face detection, and hand tracking, optimizing access control in smart door systems. Integrating software and hardware set a new standard in the domain of smart door access. Here, we have proposed a smart system with a high-resolution camera to capture faces and connect them to a processing unit, i.e., Raspberry Pi. A face detection algorithm has been deployed into the hardware device so that the lock would open once an authorized face is detected. Our proposed system can make accurate predictions compared to existing systems. Integrating advanced computer vision and authentication methods marks a significant stride in the evolution of intelligent entry solutions.

The paper presents the music therapy approach on controlling and transforming the human emotions. The proposed approach considers six basic emotions such as happy, sad, angry, fear, depression, and surprise along with neutral expression. Here, happy, surprise, and neutral are considered as positive emotions and sad, angry, fear, and depression are considered as negative emotions. Initially, the proposed approach used FER-(2013) dataset for training Convolution Neural Network (CNN) for facial expression recognition. Later, the proposed approach uses real-time video clips to analyze the emotions. In the first stage, the zeroth frame of the video clip is considered and preprocessed. Later the image frames are proceded for CNN for feature extraction and facial expression classification. For the negative emotions related facial expresssions, related category of music is played for certain time. It is noticed that facial deformation with respect to features such as eyeeyebrows, nose, and lips will change that exhibits emotion transformation. These changes are analyzed, and when music stops, the proposed approach extracts (n + T)th frame and proceeded for deep learning using CNN for facial expression classification. The experiments are performed and noticed that the proposed approach of music therapy has greater impact on transforming internal state of human emotions and supports in avoiding the future disasters. Keywords Music • Negative emotions • Facial expression • Classification • CNN 1 Introduction Nowadays, the younger generation is more subjected to negative emotions such as depression, sad, and angry. There are various reasons behind these emotional changes

Biometrics focuses on simulate the human ability to associate one or a set of corporal features of a person in a unique way by uses a specific representation, this representation is knows as identity. Visible spectrum face recognition is the identification way more natural which has a higher universality, collectability and acceptability front to the other biometrics modalities, but is weak in the invariance and distinctiveness criterion to be a good biometric. In order to improve the face recognition, the infrared spectrum arises as a good representation to solve these drawbacks in biometric identification. Buddharaju et al., proposed a process by which the vascular net is detected [4] . However, Wu et al., criticized this approach by not take into account the heat transfer between the environment and the person in the time to take the image and proposed a modification of image thermogram and show that it is a better solution to make up for the heat change [19, 21]. This paper is w...

Most of the existing systems are highly complex in terms of time and storage for recognizing user behavior. This paper proposes an emotion based recommendation system for various applications that learns the emotions of the user from the camera. The human face acts as an important part in an individual’s body and helps in detection of one’s mood. The extracted input undergoes face detection, feature extraction, emotion recognition and application identification. The application includes playing music, videos, providing motivational quotes. This eliminates the time consuming and tedious task of manually segregating a playlist or searching for motivational quotes and videos. The experimental results are obtained by classifying the emotions using convolutional neural networks and support vector machine. The data are validated using testing and training method to obtain the exact emotion of the user. The applications are identified from the predicted emotions. By this way the emotions a...

Most of the face recognition algorithms concentrate on the transformations (like DCT, FFT, etc.) for recognition of face images. These transformations concentrate on the global information of the face images and they miss the local information i.e. the relationship with the neighboring pixels. So, here, we considered the local matching method (local binary pattern) for considering the local information of the face images. For better recognition is obtained by combining the local as well as global information of the face image. So, for effective face recognition system, we combined the local binary patterns with the DCT. In this paper, we proposed a face recognition system with local binary pattern with DCT using doubly truncated multivariate Gaussian mixture model. By using EM algorithm with K-means or hierarchical clustering, the model parameters are estimated. The experimentation is carried with two face image databases, namely, Jawaharlal Nehru Technological University Kakinada (JNTUK) and Yale. The proposed system was found to be efficient compared to the existing system using GMM. The effect of the number of DCT coefficients on the recognition rate is also studied and found efficient recognition rate for 15 DCT coefficients. We also studied the recognition rate by varying the number of training images for each person.

This study details the design, development, and deployment of an Androidbased Biometric Fingerprint system tailored for institutional access control, attendance tracking, exam monitoring, and staff management. Developed collaboratively by the Innovation Centre and departments across engineering and information and communication technology (ICT), the system integrates custom hardware and software. Hardware includes fingerprint sensors connected to an ATMEGA8 microcontroller and Android interfaces for portability. The software uses modular architecture, comprising a Kotlinbased mobile app with Jetpack Compose, a Laravel-powered web admin panel, and a secure backend API hosted on a virtual private server (VPS). Fingerprint data is safely stored using base64 encoding, enabling accurate user authentication and real-time tracking. A functional prototype was built, tested, and refined, with 95 units deployed in a pilot phase. The system supports multiple fingerprint profiles, secure data handling, and integration with existing institutional platforms. Emphasizing customization, modularity, and adherence to ICT policies, the research also serves as a training tool for staff and students, enhancing operational efficiency and supporting local technology development. Performance evaluation showed a FAR of 0.5%, FRR of 1.2%, and an average authentication time of 2.3 seconds. Post-deployment, student attendance increased by 15%, fee compliance by 10%, and 89% of users rated the system as easy to use. This work demonstrates effective hardware-software co-design for scalable biometric authentication in educational settings.