Face recognition for video surveillance with aligned facial landmarks learning

Abstract

BACKGROUND:Video-based face recognition has attracted much attention owning to its wide range of applications such as video surveillance. There are various approaches for facial feature extraction. Feature vectors extracted by these approaches tend to have large dimension and may include redundant information for face representation, which limits the application of methods with high accuracy such as machine learning.OBJECTIVE:Facial landmarks represent the intrinsic characteristics of human face, which can be utilized to decrease redundant information and reduce the computation complexity. But feature points extracted in each frame of a video are irregular which needed to be aligned.METHODS:This paper presents a novel method which is based on facial landmarks and machine learning. We proposed a method to align the feature data into a common co-ordinate frame, and use a robust AdaBoost algorithm for classification.RESULTS:Experiments on the public Honda/UCSD database demonstrate the superior performance of our method to several state-of-the-art approaches. Experiments on Yale database show the sensitivity and specificity of the proposed method.CONCLUSION: The proposed methods can improve the image-set based recognition performance.

Improving Face Recognition Using Artistic Interpretations of Prominent Features: Leveraging Caricatures in Modern Surveillance Systems

“…That means checking every caricature of resemblance to the target identity and art style. Additionally, many state-of-the-art surveillance systems are reliant on facial landmarks, which can already be inaccurate in normal photos [69,70]; this inaccuracy is exacerbated in caricatures, particularly caricatures with a high degree of exaggeration across internal facial features.…”

(Retracted) Face recognition technology in classroom environment based on ResNet neural network

“…The method proposed by Lin J can improve the recognition performance based on image sets. 7 Although, the above research can promote the development of face recognition technology to a certain extent, it has not been promoted because of the disadvantages of slow calculation efficiency and high development cost. And the content of this article has gathered the characteristics of high calculation accuracy, high efficiency, and low cost.…”

Analysis of Varicose Veins of Lower Extremities Based on Vascular Endothelial Cell Inflammation Images and Multi-Scale Deep Learning

“…Literature [36] shows that smaller convolution kernels can improve predictions. Therefore, the network constructed in this chapter uses a convolution kernel with a fixed size of 3 × 3, and uses the literature [37] to construct a deep network, using a 1 × 1 convolution layer to simplify the network weight parameters, and the network training constructed. The parameter is 5650K for small-embedded devices.…”

“…Since the ReLU activation function only performs a simple linear transformation, the nonlinear representation ability is low. In literature [37], by modifying the max out activation function, an MFM activation function is proposed, which has the function of making features compact and reducing parameters. Relative to ReLU forced dilution; it retains the maximum information of the feature through a competitive mechanism, as shown in Figure 10.…”

“…The multi-scale deep learning network constructed in this paper consists of an inception model, four 3 × 3 convolutional layers, four 1 × 1 convolution kernels, nine MFM activation layers and two fully connected layers. The network integrates Google-Net and literature [37] Network structure characteristics. The detailed structure configuration is shown in Table 1.…”