Home About Login Current Archives Announcements Editorial Board
Submit Now For Authors Call for Submissions Statistics Contact
Home > Archives > Volume 20, No 13 (2022) > Article

DOI: 10.14704/nq.2022.20.13.NQ88234

Multi-view Face Recognition Using Novel Convolutional Neural Network-based Deep Learning Architecture

Bharath J, Dr R Balakrishna


Face recognition is an important research arena in the domain of human-computer interaction. For multi-view face recognition, algorithms resulting in more accurate outputs often pose a challenging bar. In this paper, a novel convolutional neural network (CNN) has been proposed and applied to multi-view facial recognition. In this proposed method, a large dataset of images is pre-processed, and raw data is altered as features. Different angles of the same image have been analysed with a deep learning model to overcome the clarity issues in the case of a multi-view domain. A CNN algorithm was applied to train the model with this complex dataset to gain an optimal output that has higher accuracy. With this applied dense layer of CNN model- training, the favourable features well suited to classify the data into separate categories were learned. The novel architecture of the model was developed using eight convolutional layers, four max-pooling layers, and a dropout layer. The Adam optimizer was used for the optimization of the performance of the proposed model. This novel CNN approach executed lower false positives and false negative outputs, yielding a high accuracy of this model. Hence, the proposed model can be used for efficient multi-view face recognition.


Convolutional neural network (CNN), multi-view face recognition, false positive, false negative, confusion matrix

Full Text