The problem of building a face recognition pipeline faces numerous challenges such as changes in lighting, pose, and facial expressions. The main stages of the pipeline include detection, alignment, feature extraction, and face representation. Each of these stages is critically important for achieving accurate recognition.
With rapid development of machine learning and subsequently deep learning, deep neural networks achieved remarkable results in solving various tasks. However, with increasing the accuracy of trained models, new architectures of neural networks present new challenges as they require significant amount of computing power for training and inference. This paper aims to review existing approaches to reducing computational power and training time of the neural network, evaluate and improve one of existing pruning methods for a face detection model.
The article is dedicated to the investigation of face identification methods and aims to determine the most suitable one for a security system based on facial recognition from surveillance cameras. The time costs of these methods and their robustness against geometric scale distortions and rotations in various planes have been analyzed. Custom datasets have been generated for experimentation purposes.