University of Malta researchers to improve the quality of facial images using Artificial Intelligence
Several countries around the world use CCTV videos as forensic evidence to combat crime. These cameras cover large fields of view, where low-resolution facial images are typically captured, making the identification of the subject of interest very difficult.
Deep-FIR has been granted around €200,000 by the Malta Council for Science & Technology, for and on behalf the Foundation for Science and Technology, through the Fusion: R&I Technology Development Programme.
This project, led by Dr Reuben Farrugia from the Department of Communications and Computer Engineering together with Prof. Kenneth Camilleri from the Depart- ment of Systems and Control Engineering at the University of Malta will adopt advanced artificial intelligence techniques based on deep learning to improve the performance of existing video forensic software packages adopted by forensic scientists in their labs.
Apart from improving the quality of facial images captured by CCTV cameras, this project is intended to reduce both the manual work of the operator and the computational complexity needed to restore a frame.
Ascent Software, a local premier software development house with a vast experience in applied AI in various ICT related sectors, is actively involved in the Deep-FIR project and will develop the first working prototype that will be evaluated on real-world cases.
An experienced researcher (postdoc) and a software developer will be engaged during this project to work on the enhancement of very low-resolution facial images and the implementation of the software package.
Apart from the members of the consortium, this project is supported by Amped Software, which is one of the leading international companies in forensic video enhancement software. Moreover, several police forces around Europe are helping the researchers define the software specifications in order to ensure that the software has the functionality required by their digital forensics teams.