🔬 Research and Innovation

Extended Family System and Welfare Care in Post COVID Nigeria: A Case Study of Warri, Nigeria

Ongoing research

This study examines how the extended family welfare system currently stands in post COVID Nigeria, focusing on intergenerational care flow patterns and the economic factors shaping them. It analyses care flows including downward support, reversed upward support, sandwich care, and horizontal support within extended families in Warri.

Data collection uses a structured household survey deployed via KoboToolbox, and planned analyses include descriptive statistics, cross tabulations with chi square tests, and multivariate regression models.

Problem: Changing family welfare responsibilities after COVID and rising economic pressure.

Method: Household survey plus statistical testing and regression modelling.

Impact: Evidence to inform social policy, welfare support, and household resilience planning.

Authors and affiliations: Chimezie Anajama (University of Sussex Business School), Warrence Oghenevwegba (Independent), Ojonugwa Wada (University of Roehampton).
Status: In progress.

Transformer Enhanced Aerial Object Detection Using YOLO and Attention Fusion

In progress

This research explores transformer based attention fusion integrated into a YOLO pipeline for aerial object detection. The work includes model design, training, evaluation, and ablation analysis across core detection metrics.

Problem: Detection accuracy drops in aerial imagery due to scale, density, and background complexity.

Method: YOLO pipeline enhanced with transformer attention fusion and evaluation across detection metrics.

Impact: Stronger detection performance for monitoring and mapping applications.

Affiliation: University of Roehampton, London, UK
Status: In progress.

Predicting the Adoption of Digital Innovations in Rural Communities: A Machine Learning Approach with Explainable AI

In progress

Digital innovations including mobile financial services, internet connectivity, and digital payment platforms can transform rural communities, yet adoption remains uneven due to differences in digital literacy, socio demographics, connectivity, and economic conditions.

This study predicts adoption outcomes using supervised machine learning and applies explainable AI methods to interpret the most influential factors, supporting targeted interventions for digital inclusion and rural empowerment.

Problem: Uneven adoption of digital services in rural communities despite rapid expansion.

Method: ML prediction models plus explainable AI interpretation of adoption drivers.

Impact: Practical insights for policy and development programmes improving financial inclusion and connectivity.