Lorna Mugambi | Google PhD Fellow

Bridging the gap between AI theory and African reality.

I am a Google PhD Fellow and Research Fellow at the Centre for Data Science and Artificial Intelligence (DSAIL), Nyeri, Kenya. My research focuses on self-supervised learning, computer vision, medical image analysis, and ecological AI.

I recently completed my MSc in Telecommunication Engineering at Dedan Kimathi University of Technology, supervised by Prof. Ciira wa Maina and Prof. Liesl Zühlke.
I'm currently a PhD student at Dedan Kimathi University of Technology (Electrical and Electronic Engineering).

Recent Awards

Oct 2025

Google PhD Fellowship — AI for Health
Aug 2025

Deep Learning Indaba 2025 — Poster Award
Apr 2025

AI Mathematical Olympiad — Progress Prize 2
Sep 2024

PhysioNet Challenge (DSA Prize) — 1st Place
Jul 2024

ACVSS 2024 Hackathon — 1st Place, Research Impact Track

Research Focus Areas

AI for Diagnostics

Developing automated diagnosis tools for Rheumatic Heart Disease (RHD). Utilizing self-supervised learning techniques to extract features from limited, unlabelled echocardiograms.

Ecological AI

Building autonomous monitoring systems for Kenyan conservancies. My work included curating and publishing the DSAIL-Porini dataset and deploying low-cost edge sensors for camera traps.

Edge Deployment

Solving the "last mile" problem. Mapping clinical user journeys to deploy large models on edge devices in regions with unreliable power grids and bandwidth.

Updates

2025

Judge — IEEE TechX Nyeri 2025

Served as a judge evaluating student projects during IEEE TechX Nyeri 2025.

Judge IEEE

Aug 2025

Deep Learning Indaba 2025

Attended Deep Learning Indaba 2025 in Kigali, Rwanda. Presented a poster on my work on self-supervised learning for rheumatic heart disease diagnosis.

Poster Award

Jun–Jul 2025

Leeds Africa Hub Visit

Attended and presented at the postgraduate conference hosted by the School of Mathematics and gave seminars to the School of Biological Sciences (University of Leeds) and the ecology department (University of Swansea).

2024

Jul–Aug 2024

Deep Learning IndabaX Uganda 2024 — Ecology Workshop Speaker

Spoke about leveraging AI for conservation preservation in Kampala, Uganda.

Jul 2024

ACVSS 2024

Ten intensive days covering dataset construction, advanced vision, and ethics-ecology. Our team won the Research Impact Track!

Jun 2024

Data Science Africa 2024 — Nyeri

Co-organized Computer Vision and Generative AI sessions covering Roboflow annotation and YOLOv8 for conservation and health.

Publications

2025

Deployment and Evaluation of the Autonomous Monitoring of Insects (AMI) System

IST-Africa Conference, 2025

G. Kiarie, L. Mugambi, J. Kabi, C. wa Maina

Developed and deployed an autonomous insect monitoring system at Dedan Kimathi University Wildlife Conservancy. The system uses image processing to count insects (avg 10 per night), providing a scalable tool for ecosystem health assessment.

Paper

Self-Supervised Multi-Task Learning for the Detection and Classification of RHD-Induced Valvular Pathology

Journal of Imaging, 2025

L. Mugambi, C. wa Maina, L. Zühlke

A comparative analysis of DINOv2 and SimCLR for RHD diagnosis. DINOv2 achieved 99% accuracy in view classification and 98% in condition detection, demonstrating the power of SSL for medical imaging with limited labels.

Paper

2024

Automated ECG Image Classification with InceptionV3

Computing in Cardiology (CinC), 2024

A. M. Gitau, V. Ruto, Y. Njathi, L. Mugambi, et al.

Fine-tuned InceptionV3 on PTB-XL dataset to classify ECG images, securing 6th place in the George B. Moody PhysioNet Challenge 2024. Addressed challenges with real-world, deteriorated paper ECGs.

Paper

The DSAIL-GeJuSTA Data Science Education Workshop: Designing a Data Science Curriculum for the African Continent

Workshop Report: Research Ideas and Outcomes, October 2024

L. Mugambi, G. Kiarie, J. Kabi, C. wa Maina, S. Mazumdar

A joint initiative by DSAIL and Gender Justice in STEM Research in Africa (GeJUSTA) to discuss designing data-science curriculum, strategies for achieving gender equity in data-science education, addressing new technological challenges in education and fostering multidisciplinary approaches to data-science education.

Paper

2023

The use of Open-Source Boards for Data Collection and Machine Learning in Remote Deployments

IEEE AFRICON

G. Kiarie, J. Kabi, L. Mugambi, C. wa Maina

Analyzed low-cost, low-power open-source boards (RPi, OpenMV) for off-grid data collection, optimizing for power consumption and processing capabilities in remote environments.

Paper

Efficient Camera Trap Image Annotation Using YOLOv5

IEEE AFRICON

Y. Njathi, L. Wanjiku, L. Mugambi, et al.

Case study using YOLOv5 to auto-annotate wildlife images. The model achieved human-level accuracy on 72% of the dataset, significantly reducing manual effort for conservationists.

Paper

Analysis of women representation in STEM in Africa

7th DeKUT International Conference on Science, Technology, Innovation and Entrepreneurship, November 2023

G. Kiarie, L. Mugambi, J. Kabi, C. wa Maina

A study analyzing the representation of women in STEM in Africa by examining the genders of members of staff in STEM faculties from African universities, analyzing the genders of STEM-papers' authors from African universities, and conducting literature review to evaluate existing measures.

Paper

2022

Towards AI Based Diagnosis of Rheumatic Heart Disease: Data Annotation and View Classification

IST-Africa Conference, May 2022

L. Mugambi, C. wa Maina, L. Zühlke

Presented a web application for labeling echocardiography data to build datasets for ML models. A foundational step towards automating RHD screening in developing countries.

Paper

DSAIL-Porini: Annotated camera trap images of wildlife species from a conservancy in Kenya

Mendeley Data, March 2022

L. Mugambi, G. Kiarie, J. Kabi, C. wa Maina

A dataset of camera trap images of wildlife species from a conservancy in Kenya with annotations. The camera traps were deployed from June 2021 to December 2021, capturing 6 categories of grazing mammals: Burchell's zebra, Defassa waterbuck, bushbuck, Common warthog, impala and the Syke's monkey.

Dataset

For a complete list of publications, visit my

Google Scholar Profile

Tutorials

Developing a Low-Cost Raspberry Pi-based Camera Trap for Wildlife Detection

DSAIL-Tech4Wildlife Workshop - November 2023

A hands-on tutorial session covering the hardware setup, software configuration, and field deployment of low-cost camera traps for conservation monitoring.

Professional & Community Engagement

• Data Science Africa 2021-2024: Speaker, organising/hosting, social media management (X), peer review
• Deep Learning Indaba, Deep Learning IndabaX-Kenya, Deep Learning IndabaX-Uganda
• Dedan Kimathi University/DSAIL Undergraduate Outreach, Tutorials & Workshop
• IEEE – Keynote speaker, Conferences, Publishing & member

Outreach

Mentoring young women in STEM.
Organizing tech workshops for underserved communities.
Collaborating with local organizations (WILDLABS Women in Conservation Technology).