Deep Learning Architectures for Face Recognition in Video Surveillance

Abstract

Face recognition (FR) systems for video surveillance (VS) applications attempt to accurately detect the presence of target individuals over a distributed network of cameras. In video-based FR systems, facial models of target individuals are designed a priori during enrollment using a limited number of reference still images or video data. These facial models are not typically representative of faces being observed during operations due to large variations in illumination, pose, scale, occlusion, blur, and to camera inter-operability. Specifically, in still-to-video FR application, a single high-quality reference still image captured with still camera under controlled conditions is employed to generate a facial model to be matched later against lower-quality faces captured with video cameras under uncontrolled conditions. Current video-based FR systems can perform well on controlled scenarios, while their performance is not satisfactory in uncontrolled scenarios mainly because of the differences between the source (enrollment) and the target (operational) domains. Most of the efforts in this area have been toward the design of robust video-based FR systems in unconstrained surveillance environments. This chapter presents an overview of recent advances in still-to-video FR scenario through deep convolutional neural networks (CNNs). In particular, deep learning architectures proposed in the literature based on triplet-loss function (e.g., cross-correlation matching CNN, trunk-branch ensemble CNN and HaarNet) and supervised autoencoders (e.g., canonical face representation CNN) are reviewed and compared in terms of accuracy and computational complexity.

Face Recognition on Low Resolution Face Image With TBE-CNN Architecture

“…And this layer can be seen as a convolution with a kernel that covers the entire input region [18]. TBE-CNN is an one of CNN architecture for extracting complementary features and patches around facial landmarks through trunk and branch networks [5]. Trunk networks contain several layers to absorb global information and branch networks contain several layers to absorb local information thereby reducing computation and effective convergence.…”

“…There are 2 ASTESJ ISSN: 2415-6698 stages of learning on CNN, namely feed-forward and backpropagation. In [5] the authors describe the four CNN architectures that are currently developi and be the subject of their research. The four architectures are Cross Correlation Mechanism CNN (CCM-CNN), TBE-CNN, HaarNet, and Canonical Face Representation CNN (CFR-CNN).…”

Revisiting crowd behaviour analysis through deep learning: Taxonomy, anomaly detection, crowd emotions, datasets, opportunities and prospects

“…In this context, automatic video-surveillance [1] emerges as a research field, attracting the attention of a large community. Areas such as dangerous object detection [2] or face recognition and identification [3] have been broadly studied.…”

EPYNET: Efficient Pyramidal Network for Clothing Segmentation

“…In recent years, Deep Learning (DL) methods have been successfully used to learn compact and discriminative features for many complex problems. In the soft biometrics area, DL methods have been used for a wide variety of classification problems, such as body-based features (height, shoulder width, hips-width, arms-length, body complexion, and hair color) [7], facial traits (gender, ethnicity, age, glasses, beard, and mustache) [8], [9], tattoos [10], gait-based gender [11], and clothes [12]- [14].…”

“…The experimental results and their discussion are shown in Section IV. Finally, general conclusions and suggestions for future research directions are presented in Section V. DL methods have been utilized to address many problems in the surveillance environment, including person reidentification [24], face recognition [9], anomaly detection [25], multi-view analysis [26], and traffic monitoring [27].…”