15 Dissertation Topics on Facial Recognition
Facial recognition technologies are increasingly penetrating our everyday life, having become a popular method of authentication for smart devices. However, the use of this authentication method has also introduced security challenges as well, all of which could potentially hinder its rapid development and implementation. There are possible legal challenges related to this phenomenon are connected to the fact that facial recognition promotes remote biometric identification. However, it is important to ascertain whether such identification has been consented upon and if so, are there circumstances where identification is justified even without consent. We have compiled a list of 15 dissertation topics looking at different perspectives of facial recognition technology, focusing on security, technology and law.
1) Biometrics for Internet of Things (IoT): An evaluation of the potential security challenges of adopting facial recognition technology for Smart Devices.
As business organisations search for improved authentication controls for their systems and networks, new security controls like facial recognition technology have also been introduced to improve existing controls. However facial recognition technology also raises concerns about potential unauthorised access by attackers who can use users’ images to access secured corporate information. The aim of this dissertation is to critically evaluate the potential challenges of facial recognition technology for IoT devices such as smartphones and personal computers, particularly with regards to access controls. Subsequently, the dissertation will propose new approaches to overcome unauthorised access control challenges for this technology.
Key Source: Jiang, R., Al-maadeed, S., Bouridane, A., Crookes, D. and Beghdadi, A. (2017). Biometric Security and Privacy: Opportunities & Challenges in The Big Data Era. Switzerland: Springer International Publishing.
2) A Comparative Study of the Challenges, Opportunities and Benefits of Fingerprints, Voice recognition, and Facial recognition technology: Which authentication method is most secure?
Though business organisations and individuals are adopting the wide range of biometric authentication methods, there appears to be a lack of understanding about the strengths and limitations of each of these methods. This dissertation aims to carry out a comparative analysis of three authentication methods, namely fingerprint, voice recognition, and facial recognition technology. The dissertation will highlight the strengths and vulnerabilities of each of these methods, and subsequently propose an approach for consolidating these authentication methods into two or three-factor authentication schemes.
Key Source: Mitra, S. and Gofman, M. (2017). Biometrics in a Data Driven World: Trends, Technologies, and Challenges. Florida, U.S.: Taylor & Francis Group, LLC.
3) Performance Evaluation: The use of the Diffusion of Innovation Model to explore the impact of facial recognition technology on users’ perception about data privacy.
Recent advancements in biometric authentication has resulted in the introduction of new authentication measures such as iPhone’s facial recognition and Touch IDs. However, there could also be resistance to users’ acceptance of this technology due to concerns about the privacy or security of their information. This dissertation aims to apply the Diffusion of Innovation theory to evaluate how users view this new technology, and if there are any users’ resistance due to concerns about privacy of personal information.
Key Source: Tryfonas, T. (2017). Human Aspects of Information Security, Privacy and Trust. Switzerland: Springer International Publishing.
4) Interoperability of Biometrics and E-commerce: The application of facial recognition technology as authentication measures by online businesses to conduct e-commerce activities.
Digital businesses face the constant challenge of providing secured and convenient e-commerce platforms for their customers. This also applies to other organisations such as financial institutions, whose daily operations are based on providing secured access to customers’ data. This dissertation aims to evaluate what role biometric authentication plays in such e-commerce transactions, and how facial recognition can be used to provide enhanced, secured payment options to customers of such businesses.
Key Source: Das, R. (2017). Adopting Biometric Technology: Challenges and Solutions. NW: Taylor & Francis Group, LLC.
5) Proposed Privacy- enhancing measures for biometric authentication (particularly facial recognition) on mobile phones.
Within biometric systems, authentication activities such as facial recognition are carried out using scoring rules that can be configured by designers. Such scoring rules can be designed to prevent false positives or false images from being authenticated to access devices using biometrics. The aim of this dissertation is to highlight the vulnerabilities facial recognition faces based on scoring rules of biometric authentication and propose a set of scoring rules that can be used to further enhance the privacy of biometric-based devices and decrease the potential for false positives on such devices.
Key Source: Brinke, D. J. and Laanpere, M. (2017). Technology Enhanced Assessment. Switzerland: Springer International Publishing.
6) The threat to privacy from facial recognition technology
Facial recognition technologies are arguably breaching our privacy. This topic should provide a discussion of the possible issues connected to this, such as breach of confidence and breach of privacy as well as the possible defences such as the interests of society. The aim should be to establish whether a balance could be stricken between the two conflicting aims of the law of privacy where facial recognition technologies are concerned.
Ursula Smartt, Media and Entertainment Law (2017, 3rd ed. Routledge)
7) Facial recognition technologies in the new Data Protection Environment
This topic aims to establish whether facial recognition technologies are or could be consistent with the GDPR. Above all, it will be looked at whether the lack of consent could be an issue, particularly when the technology is used in the context of financial services applications dealing with banking or investing. Pay by smile and other applications will be used as a starting point of the discussion.
Thuy Ong, ‘KFC in China tests letting people pay by smiling’ (4 September 2017), https://www.theverge.com/2017/9/4/16251304/kfc-china-alipay-ant-financial-smile-to-pay accessed 20th April 2018. See also, Jon Russell, ‘Alibaba debuts ‘smile to pay’ facial recognition payments at KFC in China’ (4 September 2017) https://techcrunch.com/2017/09/03/alibaba-debuts-smile-to-pay/ accessed 20th April 2018.
8) Consent related issues derived from the use of facial recognition technologies
There are a number of consent related issues, attached to the above, which need to be examined. This will require a thorough look at the GDPR and any other relevant legislation to identify the exact areas in which the lack of consent may become problematic. It will be examined whether the law provides sufficient safeguards in this context.
Intersoft consulting, ‘Art. 7 GDPR Conditions for consent’ (2018) https://gdpr-info.eu/art-7-gdpr/ accessed 20th April 2018. See also, Taylor Wessing, ‘Understanding consent under the GDPR’ (November 2016) https://united-kingdom.taylorwessing.com/globaldatahub/article-understanding-consent-under-the-gdpr.html accessed 20th April 2018.
9) Facial recognition technologies – a brave new world?
This dissertation will argue that the facial recognition technologies take us on a journey which will diminish our freedoms. It will further argue that technology has the potential of infringing on people’s privacy to the extent that the state will become intolerant and authoritarian. The discussion will explore a more philosophical aspect of the challenges to privacy presented by facial recognition technologies.
David Friedman, Future Imperfect: Technology and Freedom in an Uncertain World (2011 CUP)
10) Facial recognition technologies in the context of naming and shaming in the media
The loss of anonymity related to the use of facial recognition technologies and its legal implications should be explored. For example, there are a number of apps which allow people who transgress some rules to be identified and their identities shared on social media with the idea of shaming them into a better behaviour.
Don Pember, Mass Media Law 19th Edition 2014. Privacy in a Digital, Networked World , Sherali Zeadally and Mohamad Badra ed. (2015 Springer)
11) Facial recognition technologies on smart devices
Facial recognition-related constraints have long been researched in the field of Computer Science, and many technologies have also been implemented in the industry. Considering that face is a unique characteristic of each individual, facial recognition-related technologies were mostly implemented for surveillance (i.e. they were built based on detected facial features to allow law enforcement to query for CCTV with specific facial attributes), and recently for entertainment purposes. Moreover, through facial recognition people can use an application on their smart device to take photos and upload them on social media. This dissertation aims to give an overview about the principles behind these technologies on smart devices (such as algorithms for face detection, algorithms for face recognition, face technologies and application on smart platforms, facial feature tracking on smart platforms, etc.).
Reference: Liu, H. (2015). Face Technologies on Mobile Devices. In: Facial Detection and Recognition on Mobile Devices, 2015, 11-38.
12) Vigorous facial recognition technology using Hierarchical Collaborative Representation (HCR)
HCR is attracting the attention of scientists as it is more robust than traditional representation classifiers in facial recognition activities. For example, HCR has evidently demonstrated high facial recognition accuracy. This dissertation aims to prove that the accuracy of collaborative representation can be enhanced substantially by reducing the Euclidean distance between a face and its approximator and the Euclidian distances from the approximator to faces in each class. The recommended HCR applies a two-stage classifier for training faces to prove that the recognition accuracy of the second-stage classifier can be significantly improved compared to that of the first-stage classifier. Among others, the dissertation aims to prove that the recognition rate of the proposed classifier can be increased using models of discriminative feature extraction. Lastly, after demonstrating a number of experiments with face recognition datasets, the dissertation findings aim to demonstrate that the HCR classifier used with the models can significantly outperform the conventional state-of-the-art methodologies.
Reference: Liu, B-D., et al. (2016). Face recognition using class specific dictionary learning for sparse representation and collaborative representation. Neurocomputing 204 (2016), 198-210.
13) Technology augmentation for facial recognition
This dissertation aims to research several technology augmentation methodologies, dedicated to facial recognition to enlarge the training dataset, alleviating the impacts of pose variance, misalignment, partial occlusions, illumination changes, and the overfitting during training. The progress of each technology augmentation methodology will be based on existing testing previously conducted on several databases. The experimental findings of this dissertation aim to show that the proposed approach can significantly enhance the overall facial recognition performance.
Reference: Lv, J-J., et al. (2017). Data augmentation for face recognition. Neurocomputing 230 (2017), 184-196.
14) Facial recognition technology using Brain-Computer Interfaces (BCIs)
This dissertation is based on the research that has been conducted on the brain’s reaction to seeing reaction to facial recognitions. The unaware facial recognition is a field where additional research can be conducted and contributed to. Hence, this dissertation aims to critically review the literature to identify current experiments where participants can view images of faces while they record their electroencephalography signals using a BCI headset. The literature analysis intends to identify a number of differences between the different types of facial recognition, particularly, unaware recognitions.
Reference: Abhang, P.A., et al. (2016). Chapter 8: Brain-Computer Interface Systems and Their Applications. In: Introduction to EEG-and Speech-Based Emotion Recognition, 2016, 165-177.
15) Monitoring systems using facial recognition technology
Facial recognition is a critical branch of biometric verification; therefore, it has been widely used in several applications such as Human-Computer Interaction (HCI), Network Security, Video Monitor Systems and Door Control Systems. Considering that authentication is an important issue in system monitoring in computer based communication, this dissertation aims to propose a methodology for monitoring systems which will integrate with the facial recognition technology using the Personal Component Analysis (PCA) algorithm.
Reference: Vadiraj. M., et al. (2016). Face Recognition Based Attendance Monitoring System. International Journal of Emerging Research in Management & Technology 5 (5), 254-258.