Recognizing Very Small Face Images Using Convolution Neural Networks

Application of Face Recognition Method Under Deep Learning Algorithm in Embedded Systems

Deep learning, in particular Convolutional Neural Network (CNN), has achieved promising results in face recognition recently. However, it remains an open question: why CNNs work well and how to design a 'good' architecture. The existing works tend to focus on reporting CNN architectures that work well for face recognition rather than investigate the reason. In this work, we conduct an extensive evaluation of CNN-based face recognition systems (CNN-FRS) on a common ground to make our work easily reproducible. Specifically, we use public database LFW (Labeled Faces in the Wild) to train CNNs, unlike most existing CNNs trained on private databases. We propose three CNN architectures which are the first reported architectures trained using LFW data. This paper quantitatively compares the architectures of CNNs and evaluate the effect of different implementation choices. We identify several useful properties of CNN-FRS. For instance, the dimensionality of the learned features can be significantly reduced without adverse effect on face recognition accuracy. In addition, traditional metric learning method exploiting CNN-learned features is evaluated. Experiments show two crucial factors to good CNN-FRS performance are the fusion of multiple CNNs and metric learning. To make our work reproducible, source code and models will be made publicly available.

Secure Learning using Face Recognition has gained significant attention due to its potential solution in curbing and elimination of electronic examination impersonation through enrollment, authentication and authorization. Despite the advancements, existing system often focuses on individual entities, void of models and special dataset to enhance security measure in comprehensively curbing electronic examination impersonation. This study aims to address this gap by performing a model using face recognition: Labeled Face in the Wild (LFW), Custom Dataset, algorithms using TensorFlowJs and Face Application Programming Interface (API) model. The research evaluates this model on a unified facial recognition dataset, assessing their performance based on face verification and recognition. The LFW dataset a popular benchmark dataset for face verification and recognition focuses on detecting and alignment of captured images, where TensorFlowJs extract face embedding, while the Face API verifies captured image by loading the Custom Dataset (Enrolled images) comparing with that of the recent before successful authentication. The LFW-TensorFlowJs combines these approaches to leverage both Secure and accurate information. Results shows that integrating LFW, TensorFlowJs and Face API model outperforms the standalone LFW.

Face recognition has made extraordinary progress owing to the advancement of deep convolutional neural networks (CNNs). The central task of face recognition, including face verification and identification, involves face feature discrimination. However, the traditional softmax loss of deep CNNs usually lacks the power of discrimination. To address this problem, recently several loss functions such as center loss, large margin softmax loss, and angular softmax loss have been proposed. All these improved losses share the same idea: maximizing inter-class variance and minimizing intra-class variance. In this paper, we propose a novel loss function, namely large margin cosine loss (LMCL), to realize this idea from a different perspective. More specifically, we reformulate the softmax loss as a cosine loss by L2 normalizing both features and weight vectors to remove radial variations, based on which a cosine margin term is introduced to further maximize the decision margin in the angular space. As a result, minimum intra-class variance and maximum inter-class variance are achieved by virtue of normalization and cosine decision margin maximization. We refer to our model trained with LMCL as CosFace. Extensive experimental evaluations are conducted on the most popular public-domain face recognition datasets such as MegaFace Challenge, Youtube Faces (YTF) and Labeled Face in the Wild (LFW). We achieve the state-of-the-art performance on these benchmarks, which confirms the effectiveness of our proposed approach.

This paper describes in detail about 2D and 3D face and video face databases developed in the year from 1992 to 2009. The still face database discussed here includes: Our Laboratory Database of Faces (ORL), Yale face database, Yale face databaseB(B stands for Big), Yale face database extendedB, FERET(Face Recognition Technology) face database and CASPEAL is a large scale Chinese face database with varying PEAL(Pose Expression Accessory, and Lighting). The 3D face databases like MIT-CBCL (Centre for Biological and Computational Learning) face recognition database, Georgia Tech face database, Cohn-Kanade Action Unit Face database and 3D BFM (Basel Face Model) are also discussed in this paper. In CK AU, CK+AU(Cohn-Kanade Action Unit) coded facial expression database, a 3D model is developed from a 2D model which is obtained from the captured video of different subjects. The other Face databases and development of new Face databases are discussed at the end of this paper. The application of proposed concept is to develop a new small size face database which includes all the important criteria included in the previous face and video face databases.

Fine-grained face verification: FGLFW database, baselines, and human-DCMN partnership

Convolutional neural networks have significantly boosted the performance of face recognition in recent years due to its high capacity in learning discriminative features. In order to enhance the discriminative power of the deeply learned features, we propose a new supervision signal named marginal loss for deep face recognition. Specifically, the marginal loss simultaneously minimises the intra-class variances as well as maximises the inter-class distances by focusing on the marginal samples. With the joint supervision of softmax loss and marginal loss, we can easily train a robust CNNs to obtain more discriminative deep features. Extensive experiments on several relevant face recognition benchmarks, Labelled Faces in the Wild (LFW), YouTube Faces (YTF), Cross-Age Celebrity Dataset (CACD), Age Database (AgeDB) and MegaFace Challenge, prove the effectiveness of the proposed marginal loss.

The performance of face recognition (FR) has reached a plateau for public benchmark datasets, such as labeled faces in the wild (LFW), celebrities in frontal‐profile in the wild (CFP‐FP), and the first manually collected, in‐the‐wild age database (AgeDB), owing to the rapid advances in convolutional neural networks (CNNs). However, the effects of faces under various fine‐grained conditions on FR models have not been investigated, owing to the absence of relevant datasets. This paper analyzes their effects under different conditions and loss functions using K‐FACE, a recently introduced FR dataset with fine‐grained conditions. We propose a novel loss function called MixFace, which combines classification and metric losses. The superiority of MixFace in terms of effectiveness and robustness was experimentally demonstrated using various benchmark datasets.

Facial recognition is a critical biometric technology applied in surveillance, access control, and identity verification.However, existing Convolutional Neural Network (CNN) based models often face performance limitations under challenging conditions such as poor lighting, pose variations, occlusion, and facial expression changes.This study proposes a robust and adaptive CNN architecture to enhance recognition accuracy and generalization.The research objectives are to (i) review existing CNN based models, (ii) design an improved CNN architecture, (iii) implement and train the model using standard datasets, (iv) evaluate its performance using accuracy, precision, recall, and F1 score, and (v) compare results with baseline CNN models.The study adopts a quantitative methodology using Python based deep learning frameworks.Pre collected datasets including Labeled Faces in the Wild (LFW), CelebA, and UTKFace are processed using image normalization, face alignment via MTCNN, and data augmentation.Statistical performance metrics and confusion matrix visualization support comprehensive performance evaluation.While results demonstrate improvements, limitations include computational cost, dataset diversity, and real world deployment challenges such as latency and adaptability in dynamic environments.

In a conventional attendance monitoring system, the concerned teacher takes attendance manually in a classroom. In general, it is a time-consuming and very difficult task to take attendance of a huge number of students in a short period of time and involves proxy attendance. To overcome these issues, we proposed a face recognition-based student attendance monitoring system in a classroom environment. The proposed method uses the Histogram of Oriented Gradients (HOG) as features extractor, Convolutional Neural Network (CNN) as face encoding and Support Vector Machine (SVM) as a classifier. The proposed system recognizes the face in real-time using a webcam and generates attendance report automatically without any human intervention. Our face recognition method accomplished 99.5% accuracy on Labeled Faces in the Wild (LFW) database and 97.83% accuracy in real-time inside the classroom for the case of attendance monitoring. Finally, we tested our system to validate its effectiveness.

Chapter 10 - Local plastic surgery-based face recognition using convolutional neural networks

The intelligent campus surveillance system is beneficial to improve safety in school. Abnormal behavior recognition, a field of action recognition in computer vision, plays an essential role in intelligent surveillance systems. Computer vision has been actively applied to action recognition systems based on Convolutional Neural Networks (CNNs). However, capturing sufficient motion sequence features from videos remains a significant challenge in action recognition. This work explores the challenges of video-based abnormal behavior recognition on campus. In addition, a novel framework is established on long-range temporal video structure modeling and a global sparse uniform sampling strategy that divides a video into three segments of identical durations and uniformly samples each snippet. The proposed method incorporates a consensus of three temporal segment transformers (TST) that globally connects patches and computes self-attention with joint spatiotemporal factorization. The proposed model is developed on the newly created campus abnormal behavior recognition (CABR50) dataset, which contains 50 human abnormal action classes with an average of over 700 clips per class. Experiments show that it is feasible to implement abnormal behavior recognition on campus and that the proposed method is competitive with other peer video recognition in terms of Top-1 and Top-5 recognition accuracy. The results suggest that TST-L+ can improve campus abnormal behavior recognition, corresponding to Top-1 and Top-5 accuracy results of 83.57% and 97.16%, respectively.

With the rise of crimes associated with Automated Teller Machines (ATMs), security reinforcement by surveillance techniques has been a hot topic on the security agenda. As a result, cameras are frequently installed with ATMs, so as to capture the facial images of users. The main objective is to support follow-up criminal investigations in the event of an incident. However, in the case of miss-use, the user’s face is often occluded. Therefore, face occlusion detection has become very important to prevent crimes connected with ATM usage. Traditional approaches to solving the problem typically comprise a succession of steps: localization, segmentation, feature extraction and recognition. This paper proposes an end-to-end facial occlusion detection framework, which is robust and effective by combining region proposal algorithm and Convolutional Neural Networks (CNN). The framework utilizes a coarse-to-fine strategy, which consists of two CNNs. The first CNN detects the head element within an upper body image while the second distinguishes which facial part is occluded from the head image. In comparison with previous approaches, the usage of CNN is optimal from a system point of view as the design is based on the end-to-end principle and the model operates directly on image pixels. For evaluation purposes, a face occlusion database consisting of over 50[Formula: see text]000 images, with annotated facial parts, was used. Experimental results revealed that the proposed framework is very effective. Using the bespoke face occlusion dataset, Aleix and Robert (AR) face dataset and the Labeled Face in the Wild (LFW) database, we achieved over 85.61%, 97.58% and 100% accuracies for head detection when the Intersection over Union-section (IoU) is larger than 0.5, and 94.55%, 98.58% and 95.41% accuracies for occlusion discrimination, respectively.

Face swapping refers to the task of changing the appearance of a face appearing in an image by replacing it with the appearance of a face taken from another image, in an effort to produce an authentic-looking result. We describe a method for face swapping that does not require training on faces being swapped and can be easily applied even when face images are unconstrained and arbitrarily paired. Our method offers the following contributions: (a) Instead of tailoring systems for face segmentation, as others previously proposed, we show that a standard fully convolutional network (FCN) can achieve remarkably fast and accurate segmentation, provided that it is trained on a rich enough example set. For this purpose, we describe novel data collection and generation routines which provide challenging segmented face examples. (b) We use our segmentations for robust face swapping under unprecedented conditions, without requiring subject-specific data or training. (c) Unlike previous work, our swapping is robust enough to allow for extensive quantitative tests. To this end, we use the Labeled Faces in the Wild (LFW) benchmark and measure how intra- and inter-subject face swapping affect face recognition. We show that intra-subject swapped faces remain as recognizable as their sources, testifying to the effectiveness of our method. In line with established perceptual studies, we show that better face swapping produces less recognizable inter-subject results (see, e.g., Fig. 2.1). This is the first time this effect was quantitatively demonstrated by a machine vision method. Some of the material in this chapter previously appeared in [47].

In face recognition, face rotation alignment is an important part of the recognition process. In this paper, we present a hierarchical detector system using eye and eye-pair detectors combined with a geometrical method for calculating the in-plane angle of a face image. Two feature extraction methods, the restricted Boltzmann machine and the histogram of oriented gradients, are compared to extract feature vectors from a sliding window. Then a support vector machine is used to accurately localize the eyes. After the eye coordinates are obtained through our eye detector, the in-plane angle is estimated by calculating the arc-tangent of horizontal and vertical parts of the distance between left and right eye center points. By using this calculated in-plane angle, the face is subsequently rotationally aligned. We tested our approach on three different face datasets: IMM, Labeled Faces in the Wild (LFW) and FERET. Moreover, to compare the effect of rotational aligning on face recognition performance, we performed experiments using a face recognition method using rotationally aligned and non-aligned face images from the IMM dataset. The results show that our method calculates the in-plane rotation angle with high precision and this leads to a significant gain in face recognition performance.