From Eyes to Face Synthesis: a New Approach for Human-Centered Smart Surveillance

Xiang, Chen; Qing, Linbo; He, Xiaohai; Su, Jie; Peng, Yonghong

doi:10.1109/access.2018.2803787

Cited by 25 publications

(17 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…FacenessNet and varaints [171,173] Grid loss [109], AOFD [18], Hierarchical aƩenƟon [160] R-CNN and variants [39,40,122], SSD [87], YOLO [120], ReƟnaFace [26], FANet [181] DPM based [163] F I G U R E 4 Categorisation of methods used in occluded face detection ZENG ET AL.…”

Section: ) Locate Visible Facial Segments 3) Deep Learningmentioning

confidence: 99%

A survey of face recognition techniques under occlusion

2021

View full text Add to dashboard Cite

The limited capacity to recognise faces under occlusions is a long-standing problem that presents a unique challenge for face recognition systems and even humans. The problem regarding occlusion is less covered by research when compared to other challenges such as pose variation, different expressions, etc. Nevertheless, occluded face recognition is imperative to exploit the full potential of face recognition for real-world applications. In this article, the scope to occluded face recognition is restricted and a systematic categorisation that new as well as classic methods fit into is presented. First, the authors explore the kind of the occlusion problem and the type of inherent difficulties that can arise. As a part of this review, face detection under occlusion, a preliminary step in face recognition. Second the authors analyse how the existing face recognition methods cope with the occlusion problem and classify them into three categories, which are given as: 1) occlusion robust feature extraction approaches, 2) occlusion aware face recognition approaches, and 3) occlusion recovery based face recognition approaches. Furthermore, the motivations, innovations, pros and cons, and the performance of representative approaches for comparison are analyzed. Finally, future challenges and method trends of occluded face recognition are thoroughly discussed.This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

show abstract

Section: ) Locate Visible Facial Segments 3) Deep Learningmentioning

confidence: 99%

A survey of face recognition techniques under occlusion

2021

View full text Add to dashboard Cite

show abstract

“…To overcome and tackle this problem, many researchers have worked on the different directions of face synthesis. These directions include, converting the face edges into the natural face images [23], swapping the facial attributes of two different face images [24], generating the face with the help of the side face [25], generating the face with the help of the human eye's region [26], draw a sketches from the human face [27], face make-up [28] etc. But as per our best knowledge, no one combined the different face related information in a single methodology to generate the natural and realistic face images.…”

Section: B Text To Face Generationmentioning

confidence: 99%

A Realistic Image Generation of Face From Text Description Using the Fully Trained Generative Adversarial Networks

Khan

Saba

et al. 2021

IEEE Access

View full text Add to dashboard Cite

Text to face generation is a sub domain of text to image synthesis, and it has a huge impact along with the wide range of applications on public safety domain. Currently, due to the lack of dataset, the research work focused on the face to text generation is very limited. Most of the work for text to face generation till now based on the partially trained generative adversarial network, in which the pre-trained text encoder has been used to extract the sematic features of input sentence. Then these semantic features have been utilized to train the image decoder. But in this research work, we have proposed the fully trained generative adversarial network to generate the realistic and natural images. We have trained the text encoder as well as the image decoder at the same time to generate the more accurate and efficient results. In addition to proposed methodology, we have also generate the dataset by the amalgamation of LFW, CelebA and locally prepared dataset. We have also labelled the images according to our defined classes. Through performing different kind of experiments, we have proved that our proposed fully trained GAN outperformed by generating the good quality images with accordance to the input sentence. Moreover, the visual results have also strengthened our experiments by generating the face images according to the given query.

show abstract

“…The main idea is to operationalize and implement the system. [39] proposed the use of deep learning for smart security surveillance but was limited only on the face information and on the available data of the eyes' region. The experimental results of the research work demonstrated that the proposed method can predict the terrorist elements based on eye-data only.…”

Section: Computational Approachmentioning

confidence: 99%

Computational Robotics: An Alternative Approach for Predicting Terrorist Networks

Nwanga¹,

Okafor

Chukwudebe

et al. 2021

Int. J. Robot. Autom. Technol.

View full text Add to dashboard Cite

Increasing terrorist activities globally have attracted the attention of many researchers, policy makers and security agencies towards counterterrorism. The clandestine nature of terrorist networks have made them difficult for detection. Existing works have failed to explore computational characterization to design an efficient threat-mining surveillance system. In this paper, a computationally-aware surveillance robot that auto-generates threat information, and transmit same to the cloud-analytics engine is developed. The system offers hidden intelligence to security agencies without any form of interception by terrorist elements. A miniaturized surveillance robot with Hidden Markov Model (MSRHMM) for terrorist computational dissection is then derived. Also, the computational framework for MERHMM is discussed while showing the adjacency matrix of terrorist network as a determinant factor for its operation. The model indicates that the terrorist network have a property of symmetric adjacency matrix while the social network have both asymmetric and symmetric adjacency matrix. Similarly, the characteristic determinant of adjacency matrix as an important operator for terrorist network is computed to be -1 while that of a symmetric and an asymmetric in social network is 0 and 1 respectively. In conclusion, it was observed that the unique properties of terrorist networks such as symmetric and idempotent property conferred a special protection for the terrorist network resilience. Computational robotics is shown to have the capability of utilizing the hidden intelligence in attack prediction of terrorist elements. This concept is expected to contribute in national security challenges, defense and military intelligence.

show abstract

From Eyes to Face Synthesis: a New Approach for Human-Centered Smart Surveillance

Cited by 25 publications

References 17 publications

A survey of face recognition techniques under occlusion

A survey of face recognition techniques under occlusion

A Realistic Image Generation of Face From Text Description Using the Fully Trained Generative Adversarial Networks

Computational Robotics: An Alternative Approach for Predicting Terrorist Networks

Contact Info

Product

Resources

About