PhD Studentship [self-funded]: Modelling and classifying concealed face recognition - Jan 2021
Applications are invited for a self-funded PhD project to test and model factors that interact with correct classification of concealed face recognition. Concealed face recognition is that act of denying recognition of someone you know (e.g., perpetrators lie about knowledge of associates during criminal investigations, patients might fake amnesia to make fraudulent clinical claims). Concealed face recognition thus presents an important scientific and societal problem.
Traditional polygraph lie detection is unreliable; the critical nature of the question “Did you kill person x” is evident whether you are guilty or not. Innocent suspects may well respond more strongly than a guilty practised liar. Alternatively, the Concealed Information Test is the extensively validated for the detection of concealed knowledge (Ben-Shakhar & Elaad, 2003; Hira & Furumitsu, 2002; Meijer et al., 2014; Verschuere, Ben-Shakhar & Meijer, 2011). In Japan, approximately 5000 CITs are conducted each year using autonomic measures of recognition (e.g. skin conductance) based on knowledge of crime details. However to date, there has been little work evaluating the CIT for detecting recognition of faces (see Millen et al, 2017; Millen & Hancock, 2019; Millen, Hope & Hillstrom, 2020). Accordingly, there are unanswered questions about the reliability of implicit markers of recognition based on individual differences in memory performance, the ability to identify faces of other races, and central executive performance such as the ability to manage response conflict during lies.
This PhD project will combine the CIT with a range of methodological approaches (e.g., eye tracking, autonomic measures, EEG, neurocomputing) to clarify the relationship between cognitive processes, individual differences and implicit markers of concealed recognition for detection.
Application Requirements: Eligible applicants should – • Hold a Bachelor’s degree in Psychology, Computer Science and related fields. A relevant MSc qualification is desirable. • Be analytical and comfortable with managing large data sets including biopsychology and psychophysiology data. • Have confidence in using a statistical package such as R
Funding Notes: The PhD project is self-funded.
Tuition fees are available at: View Website
- PhD Studentship: Animal health in marine ecosystems: machine learning and data science to discover structure in unstructured data - March 2019
- PhD Studentship: Domain Specific Optimisation Techniques for Real-time Image/Signal Processing on Heterogeneous FPGA+CPU platform - March 2019
- [2019-2022] (ST Microelectronics/Univresity of Stirling: £143.75K) PhD studentship: Domain specific optimisations for real-time image processing on heterogeneous FPGA+CPU+GPU
- [2019-2023] (Scotland's Rural College/Univresity of Stirling: £79K) PhD studentship: Improving animal health in marine ecosystem using data mining and signal/image processing