Dr Shahroz Nadeem

During my time at the University of Derby, I held multiple teaching positions, including an Associate Lecturer. As the module leader for the course Network Investigation, I took great pride in delivering engaging lectures both online and face-to-face. I also took on responsibilities such as marking assignments, preparing teaching materials, providing student support, and managing administrative tasks.

Before that, I served as a Post Graduate Teaching Assistant (PGTA). In this role, I had the opportunity to work closely with lecturers in teaching various lab sessions. I was responsible for designing lab content and delivering interactive sessions for courses like Cyber-crime Analysis and Investigation, Digital Forensics, and Databases.

Prior to my time at the University of Derby, I worked as a Lecturer at the National University of Computer and Emerging Sciences. In this role, I primarily focused on designing, conducting, and evaluating programming language labs. I supported students in honing their programming skills and provided guidance on using different tools. Teaching courses such as Advance Programming (JAVA), Computer programming (C++), and Introduction to Computers (C++, Python) allowed me to create a positive and enriching learning environment.

In my professional research experience, I have worked as a Research Associate at the Cyber Security Research Group. My primary focus has been on violence detection using virtual platforms. The objective of my research was to develop synthetic virtual datasets that address ethical and moral implications related to sensitive data, particularly violence. Creating methods that can be trained on virtual data and subsequently transfer the learned knowledge to real-world applications. Throughout my research, I have extensively utilized Keras for deep learning applications, leveraging its capabilities in my work. This experience has enhanced my proficiency in implementing advanced deep learning models.

Prior to my current position, I worked as a Research Assistant at the ReVeal Lab from January 2015 to April 2018. During this time, I focused on deep convolutional neural networks for image restoration, specifically in blind textual deblurring of images. I conducted empirical evaluations to analyze the learning behavior of convolutional neural networks, exploring the effects of different design choices on their performance in image restoration tasks.

In addition to my research work, I extensively utilized Google's TensorFlow and UC Berkeley's CAFFE frameworks during this period. I also had the opportunity to conduct a week-long Data Science training boot camp in collaboration with the Higher Education Commission (HEC). Furthermore, I was offered an internship during my Final Year Project, which was centered around Openstack.

Publication list
1. Nadeem, M.S., 2021. Deep Labeller: Automatic Bounding Box Generation for Synthetic Violence Detection Datasets. Submitted to Multimedia Tools and Applications
2. Nadeem, M.S., ul Hussain, S. and Kurugollu, F., 2021. Textual Deblurring using Convolutional Neural Network. In TechRxiv
3. Nadeem, M.S. and Kurugollu, F., 2021. Effectiveness of Synthetic Images in Violence Detection. In techRxiv
4. Nadeem, M.S., Franqueira, V.N., Kurugollu, F. and Zhai, X., 2019, December. WVD: A New Synthetic Dataset for Video-Based Violence Detection. In International Conference on Innovative Techniques and Applications of Artificial Intelligence (pp. 158-164). Springer, Cham.
5. Nadeem, M.S., Franqueira, V.N. and Zhai, X., 2019. Privacy verification of PhotoDNA based on machine learning. The Institution of Engineering and Technology (IET).
6. Nadeem, M.S., Franqueira, V.N., Zhai, X. and Kurugollu, F., 2019. A survey of deep learning solutions for multimedia visual content analysis. IEEE Access, 7, pp.84003-84019.
7. Rahim, R. and Nadeem, S., 2018. End-to-end trained cnn encoder-decoder networks for image steganography. In Proceedings of the European Conference on Computer Vision (ECCV) Workshops (pp. 0-0).