Abstract:With the application of deep convolutional neural networks, the performance of computer vision tasks has been improved to a new level. The construction of a deeper and more complex network allows the face recognition algorithm to obtain a higher accuracy, However, the disadvantages of large computation and storage costs of neural networks limit the further popularization of the algorithm. To solve this problem, we have studied the unified and efficient neural network face recognition algorithm under the condition of a single camera; we propose that the complete face recognition process consists of four tasks: face detection, in vivo detection, keypoint detection, and face verification; combining the key algorithms of these four tasks, we propose a unified network model based on a deep separable convolutional structure—UFaceNet. The model uses multisource data to carry out multitask joint training and uses the keypoint detection results to aid the learning of other tasks. It further introduces the attention mechanism through feature level clipping and alignment to ensure the accuracy of the model, using the shared convolutional layer network among tasks to reduce model calculations amount and realize network acceleration. The learning goal of multi-tasking implicitly increases the amount of training data and different data distribution, making it easier to learn the characteristics with generalization. The experimental results show that the UFaceNet model is better than other models in terms of calculation amount and number of parameters with higher efficiency, and some potential areas to be used.

Layer-wise Supervised Neural Network for Face Alignment with Multi-Task Regularization

Multi-task Convolution Network for Face Alignment

Joint Learning for Face Alignment and Face Transfer with Depth Image

Heatmap-Guided Selective Feature Attention for Robust Cascaded Face Alignment

Joint Face Detection and Alignment Using Multitask Cascaded Convolutional Networks

Learning a Multi-Center Convolutional Network for Unconstrained Face Alignment.

Multitask learning and CNN for application of face recognition.

Face Verification with Multi-Task and Multi-Scale Feature Fusion

Pose-Invariant Face Alignment with a Single CNN

A Multi-Task Collaborative Learning Method Based on Auxiliary Training and Geometric Constraints

Multi-Attention Network for 2D Face Alignment in the Wild.

Deep Multi-Center Learning for Face Alignment

Multistage Model for Robust Face Alignment Using Deep Neural Networks

UFaceNet: Research on Multi-Task Face Recognition Algorithm Based on CNN.

Joint Face Detection and Alignment Using Focal Loss-Based Multi-task Convolutional Neural Networks

Branched Convolutional Neural Networks for Face Alignment

Multiscale recurrent regression networks for face alignment

Facial Landmark Detection Via Cascade Multi-Channel Convolutional Neural Network

Face Alignment by Deep Convolutional Network with Adaptive Learning Rate.

Multi-task Learning of Cascaded CNN for Facial Attribute Classification.