Skip to main content Skip to navigation

Olayinka Ajayi


Currently, I am enrolled as a student of the MathSys II CDT studying for a PhD in Mathematics of Real-world Systems. My research broadly spans the field of computer vision, graphs/network theory, video processing and machine learning. Whatever requires logic, algorithms and programming (combined) is my forte!

Particularly, my PhD research is in video representation learning using deep spatiotemporal models. Currently, I focus on human action recognition using skeletal graph data. For each human in a video frame, we extract the skeletal graph (the 3D location of selected joints on the body) with the corresponding adjacency matrix. This information is fed into the deep spatiotemporal model, which aims to identify the action performed by the human(s) in the video clip.

The proposed machine learning model can be used for downstream tasks such as surveillance, caption generation, video retrieval, and visual storytelling.

You can find a more detailed overview of my current and past research here.


PhD in Mathematics of Systems | University of Warwick

2021 - Present

MSc in Mathematics of Systems | University of Warwick

2020 - 21


MSc in Applied mathematical science with climate change impact modelling | Heriot-Watt University

2019 - 20


BSc in Mathematics and Computer Science | University of Port Harcourt

  • BSc Project: "An Implementation of the Rivest–Shamir–Adleman (RSA) Public Key Cryptosystem (PKC)".
  • Supervisor: Ugorji Chimezie

2014 - 18

First class


Conferences and Contributed Talks

  • SPAAM Seminar | University of Warwick, 8 December 2022 (and)
  • Warwick (Undergraduate) Mathematics Society Seminar | University of Warwick, 6 December 2022
    Talk: "Deep Spatiotemporal Models for Surveillance"

Have you ever wondered how surveillance is achieved by most governments of the world? How they are able to regularly and accurately monitor your physical activities? Though conversations on politics, policies and ethical correctness are relevant for this discussion, rather we would focus on the design of the machine learning models that helps achieve surveillance. In particular, I would be discussing my research on human action recognition using graph neural networks in addition to temporal models.

  • Signal and Information Processing Lab (SIPLab) Seminar | University of Warwick, 19 October 2022
    Talk: "Human Action Recognition Using Human Skeletal Graph"

In an effort to teach machines how to recognize human actions, many literatures have proposed how to achieve this using the human skeletal structure alongside the coordinates of specific joints on the human body. To make the best use of the skeletal structure, graph neural networks (GNNs) are proposed to help aggregate information across the human skeletal graph. This approach only deals with the spatial domain of the video data and leaves room for creativity in the temporal domain. The talk would focus on providing a background on the use of GNNs in human action recognition, and the approach I have adopted to get comparable results on benchmark datasets.

  • MathSys CDT Annual Conference | University of Warwick, 25 - 27 April 2022


Academic year 2022/23

Term 1

  • Associate tutor for Intro to ComputingLink opens in a new window for the 2022 MathSys MSc cohort
    • giving an overview of bash scripting, python, GitHub and an introduction to HPC.
  • TA for Matrix Analysis and Algorithms (MA398Link opens in a new window, third-year maths module)
    • solve systems of linear equations, least-squares problems, and eigenvalue problems, using highly efficient solvers.
    • understanding how to construct relevant algorithms (using python) and problems central in numerical linear algebra and to analyse them with respect to accuracy and computational cost.
    • support with marking assignments.
  • TA for Mathematics by Computer (MA124Link opens in a new window, first-year maths module)
    • introduce the rudiments of computer programming (using python) to first-year undergraduates.
    • provide computer lab support sessions for python projects/assignments.
    • support with marking assignments.

Academic year 2021/22

Term 2

  • TA for Differential Equations: Modelling and Numerics (MA261Link opens in a new window, second-year maths module)
    • help students understand the central concepts of mathematical modelling and analyse fundamental numerical methods.
    • implement and test numerical methods using python.
    • provide computer lab support sessions for python projects/assignments.
    • support with marking assignments.
  • TA for Algorithmic Graph Theory (CS254Link opens in a new window, second-year computer science module)
    • help students understand the basics of graphs as it relates to practical examples. Design effective algorithmic techniques to study basic properties of graphs and for various optimisation problems on graphs.
    • support with marking assignments.

Other Activities (including non-academic)

  • SPAAM Seminar Organizer for the Warwick SIAM Student Chapter.
2022 - 2023
2022 - 2024
2022 - present