We propose to construct a system for 3D face recognition. We propose to use photometric stereo for face reconstruction in order to by pass the problems of conventional stereo (that needs to solve the matching problem first), structured light (that does not supply colour information) and photometric stereo with spectrally distinct light sources (that relies on the assumption of uniformly coloured imaged objects). Photometric stereo (PS) can reproduce structural details and colour on a per pixel basis in a way that no other 3D system can. The proposed scheme will be appropriate for use in a controlled environment for authentication purposes, but also in a general environment e.g. the entrance of a public event. We shall use two routes: surface reconstruction from the data and direct extraction of facial characteristics from the PS set. In the first approach, once surface normal and albedo is recovered, images of the face may be synthetically rendered under arbitrary new pose and illumination conditions to allow novel viewing conditions. We also aim to use a new multi-scale facial feature matching approach in the recognition process, where facial features range from overall face and head shape to fine skin dermal topography, reflectance and texture. The latter may be thought of as a form of detailed surface bump map forming a unique skin-print or signature and represents a new approach. Hence both the 3D shape and 2D intensity data will be used in recognition or authentication tasks. We propose to use scalable methods for matching, so we can cope with large databases. 3D matching will be done with the newly proposed invaders algorithm which is FFT cross-correlation based, and more detailed matching will be done by using features and classifier combination. The novelty of our approach lies in the use of PS to extract 3D information, the use of detailed facial characteristics like moles, scratches, and skin texture, and in the design of the system so that it can operate while the person is moving, with minimum intrusion and maximum efficiency. We have two industrial collaborators who will contribute in system design, data gathering and exploitation and support from the Home Office. We shall evaluate our system following three possible scenaria: a face searched in the crowd (real time face recognition), a person has to be identified (off-line face recognition) and a person has to be checked against a claimed identity (face authentication). We shall install the first prototype system in the offices of one of our industrial partners in month 12, so that data can be collected. We envisage a door like structure with lights flashing in succession as a person walks through, while a camera is capturing images. We propose to investigate the optimal number of lights in terms of efficiency and accuracy of the reconstruction, and the option of using non-visible light to avoid problems with people sensitive to flashes. We shall also investigate the relationship between detail that has to be captured and the geometry of the construction.

The COVID-19 crisis is changing the shape of crime. Drawing on crime science, this research will inform evidence-based policy and practice. Lockdown requires people to stay home, leading to domestic violence and child abuse increases. Yet social distancing means police are arresting fewer suspects: reduced services at time of greater need. COVID-19 gives fraudsters a 'conversation starter' to approach people in-person, via text, email and online. Remote working and online leisure activities, furloughs and financial difficulties, provide more potential targets for online crimes of various types. Vulnerable groups including the elderly and disabled are more at risk. Yet a Harvard study (Kissler et al. Science, 14 April) suggests that, absent a vaccine, social distancing may continue into 2022, perhaps 2024. So we will anticipate crime effects of prolonged, graduated or cyclical exit strategies. We will also anticipate post-crisis scenarios, seeking to sustain declines in crimes like burglary, to avoid them returning to 'normal'. We will use (1) national police data, (2) detailed data from three police partners, (3) fraud and e-crime data from industry, and (4) sources from other agencies such as Childline (for unreported crime). Pre/post-change analysis will use a combination of time-series and spatial modelling. Nesting force-level analysis in the national and international context will allow us to gauge scalability. We have police and industry partners, national (Home office, National Police Chief's Council, College of Policing) and international advisors. The aim is to inform policy and practice, producing 16 deliverables including policy and practice briefings and research articles.

Gun crime is a problem in some areas in the UK and many other countries. The police have ways of detecting criminals carrying guns, but this usually involves surveillance over a period of time together with targeted stop and search. There is no affordable detector available that is capable of remotely sensing whether a person is likely to be carrying a gun or not. The police do have scanners and portals that can be installed at key locations such as airports and major event venues to detect people carrying even small metal objects, or portals that use harmless microwaves, THz waves or very low level x-rays to form images of concealed guns on a person. THz and microwaves can form clear body images by penetrating clothing, but moral and technical issues arise from the technology, particularly intrusion into privacy. Also, these are not easily deployable, however, and are still at the research stage in many cases. What is really required is a portable device capable of remotely and discretely detecting whether suspects are carrying weapons and this proposed project aims to commence the development of such a device. During its development, the research will aim to first identify what sort of electromagnetic radiation best penetrates clothing over a range of atmospheric conditions. Microwaves are a form of electromagnetic radiation, but other forms exist such as light, infra-red, Tera-Hertz and millimetre waves and they all differ in their ability to penetrate fabrics. It is also possible to use ultrasound to detect metal objects concealed under clothing and we also propose to investigate this. Some of these forms of electromagnetic radiation get absorbed by the body, whereas others are reflected back depending on the precise wavelength. We will be looking for reflections off the surfaces of the gun in a similar manner to radar detecting the bright echoes from ships and aircraft, whilst filtering out the lower level reflections from the human body. Guns are not the only objects that could be concealed about a person that could give these bright reflections at a remote sensing site. Mobile phones, leather wallets, pens and portable music players could also give detectable signals. We aim to use features unique to a gun, such as gun barrels and other cavities to identify unique signatures in the reflected signals. For example, the gun barrel acts as a resonant cavity rather like air blown over a musical wind instrument, and we aim to detect these resonances remotely. During the second phase of the investigation, we will utilise a mix of the most effective bands in the electromagnetic spectrum, whether that be radio frequencies, microwaves or some other part of the spectrum, together with ultrasound, to develop a sensor that is effective at detecting guns remotely and is deployable by the police. It is possible that different guns will produce different responses from the sensor, but we will use pattern recognition techniques similar to those used in the automatic recognition of number plates or handwriting (i.e. neural networks ) to learn to recognise these particular responses. The research will be undertaken by a consortium of Universities, namely Manchester Metropolitan University, Manchester University, University of Huddersfield and Queen Mary London who will each investigate different aspects of the problem.

Together with industrial partners, we have established that there is a strong unmet demand for individuals with expertise in the combination of statistics, applied mathematics, computation, and the collaborative problem solving skills required to acquire application area knowledge. Consider, for example, aircraft structural design, where statistical methods have recently been approved in the certification of aircraft, complementing traditional experimental testing. This ushers in a change in possible design methodology, but creates a corresponding gap for the necessary talent in the workforce: scientists with knowledge of materials, computational methods and statistics. Such individuals are needed to sustain the UK's global competitive advantage, industrially and academically. We propose a world leading and innovative cohort-driven centre for doctoral training at the interface of Statistics and Applied Mathematics: Statistical Applied Mathematics at Bath (SAMBa). Modern mathematical models describing real world applications must incorporate randomness and data in a variety of ways in order to improve their ability to predict complex behaviour and describe empirical observations. Traditionally, deterministic applied mathematics and statistical methods have taken different approaches in modelling observed phenomena. More recently, we have seen that this is proving to be a hindrance to the competitiveness of British mathematics, especially when taking account of the enormous scope for research with genuine real-world impact. SAMBa will create a new generation of interdisciplinary mathematicians, both for academic careers as well as for insertion into British industry. Their primary strengths will be their problem solving ability and their fearlessness of barriers separating mathematical modelling and modern statistics. Moreover, the implementation of this CDT will promote a novel way of educating UK PhD students within the mathematical sciences, in which there is horizontal cross-disciplinary and industrial integration through CDT activities. The central mechanism by which this horizontal integration will occur will be through week-long Integrative Think Tanks (ITT), which share similarities with sandpits. These ITTs will be supported by an array of new courses that span a spectrum including statistics, stochastic simulation and applied mathematics. SAMBa will enrol ten students per year on a four-year study programme. The first year will focus on the new courses and in the formation of research themes, as well as developing cohort integration. ITTs will occur at the end of the first and second semesters during the first year of study, and will give students the opportunity to learn how to formulate problems and structure their approach to problem solving. ITTs will be intensive activities, managed by academic staff together with interdisciplinary and industrial leaders. Students in later years will participate in one ITT per year with a view to enhancing the PhD cohort experience. The expected outcomes of the ITTs will be: to provide real experiences in approaches to problem solving, to promote cross-fertilisation of ideas and expertise through horizontal integration, to build a cohesive PhD student cohort, to catalyse new collaborations, and to provide a source of PhD thesis projects. It is expected that most, but not all, PhD thesis problems and supervisory teams will emerge from ITTs. PhD students will also run a symposium series to prepare for, and subsequently reinforce, the ITT experience as well as to develop the students' sense of research empowerment. Students in SAMBa will be awarded an M.Res. after one year, subject to successful assessment. In addition, we will strongly encourage three month industrial or cross-disciplinary academic placements. These placements will enhance the horizontal integration and are a natural extension of our long-standing and thriving BSc an MSc placement scheme.

This project will develop an operationally and technically viable approach to cargo threat investigation. The main aim of the project is to provide a real-time dynamic passive profiling technique to assist Border Control Agencies and has the potential to improve hit rates; i.e. to improve targeting the people that carry contraband and hence ensure less is entering the UK.To be specific, the real-time dynamic passive profiling technique will be based on the modelling of facial expressions, eye movement and pupil changes in both the visual and thermal domains and link these to malicious intent and physiological processes (such as blood flow, eye movement patterns, and pupil dilation). To facilitate this process, one of the initial aspects of the project will be the collection, analysis and development of the dataset used to model the baseline of facial imagery behaviour of the general population against which physiological behaviours in people with malicious intent would need to be detected. Both the baseline and the dynamic profiling will be based on the response to a series of questions. The developed techniques will be evaluated in operational trails at border control points. The multi-modal facial analysis will provide additional information to the current profiling and the developed techniques will have a wider remit into other domains. It is envisioned that this will be easily integrated into the current process.There are three main challenges:a) to determine the facial/eye features, in combination with psychological profiling, to provide robust baselines that can be linked to malicious intent,b) to develop and combine the various dynamic real-time facial models (visual expression, thermal, eye movement) related to intent, andc) to evaluate the developed system within different environments, ranging from airport to port based border control points.