Abstract <p>Deep learning algorithms, particularly convolutional neural networks, have led to advancements in various fields, including biometric authentication. Deep convolutional neural networks conduct feature extraction and classification by training the entire system in an end-to-end framework, obviating the requirement for manual feature extraction. In this paper, a multialgorithm face biometric is proposed to be developed leveraging the power of transfer learning in convolutional neural networks. Transfer learning is used to offset the dataset size limitation and save on computing resources. Face, images are first subjected to preprocessing such as augmentation and resizing. Then, features are extracted using pretrained networks: AlexNet, VGG-16 and InceptionV3. The dimensions of features are reduced using a combination of principal component analysis and linear discriminant analysis, with classification done using linear and quadratic kernels of support vector machines. The effectiveness of the proposed approach was evaluated on VIdTIMIT face dataset.</p>