Greetings from DSAIL!
It is our sincere hope that you are doing great.
Check out our newsletter featuring some of our articles, blogs and news for the past month and we hope you will enjoy reading them.

Top News and Articles

DSAIL team attended the Data Science Africa 2023 event. Data Science Africa (DSA) holds such one-week events annually featuring a summer school and workshop. The summer school, which aims to introduce students and researchers to the fundamentals of data science and machine learning takes place in the first three days; while the workshop which revolves around paper presentations and panel discussions lasts two days.
This year, the event took place from 8^th to 12^th May at the University of Rwanda in Kigali with the theme of "Harnessing Data Science for Africa's Socio-Economic Development".
Participants at DSA 2023 had the opportunity to select their preferred track, either AI or Data Science. Additionally, there was introduction of DSA Hackathons, encouraging collaboration among attendees across countries and institutions to address data science challenges in the African context. The event attracted attendees from various African countries and beyond, representing different backgrounds and institutions.
DSA Kigali was culminated with a visit to the Kigali Genocide Memorial and later a tour of the Nyandungu urban wetland Eco-tourism park.
We would like to acknowledge and thank ARM for their support, which made it possible for the DSAIL team to participate in the DSA 2023 event.

DSAIL has successfully submitted three papers for the IEEE AFRICON 2023 conference, all of which have been accepted for publication. These papers focus on data collection, annotation, and machine learning methods. Gabriel et al's paper explores the use of open-source boards for data collection and machine learning in remote deployments. Yuri et al's paper showcases the efficiency of YOLOv5 in annotating image data using the DSAIL Porini dataset . Additionally, Stephen et al's paper discusses the application of a Deep LSTM-Autoencoder for anomaly detection in sensor data, specifically using water level data collected along the River Muringato by Jason Kabi - a researcher at DSAIL to facilitate monitoring of catchment areas.The IEEE AFRICON 2023 conference is scheduled to take place in Nairobi, Kenya, in September.

A team from DSAIL presented their research papers at the IST-Africa 2023 Conference, focusing on the areas of health and enviromental conservation. In the health category, Victor et al. showcased the application of computer vision for measuring knee flexion angle to assist in monitoring orthopedic patients. Yuri et al. delved into the unsupervised discovery of echocardiographic views to aid in the diagnosis of RHD disease. In a different context, Jason et al. explored anomaly detection in IoT data, while Muhinyia et al. investigated the screening of bats' bioacoustic recordings to support ecosystem monitoring. The conference, held at the CSIR international conference centre in Tshwane, South Africa, from may 30^th to june 2^nd, provided an excellent platform for the team to present their papers. The papers were well received by the research community, offering valuable feedback to guide further research.

We would like to express our gratitude to the Kenya Education Network (KENET) for the travel grants that made it possible for our team to attend the conference.

Following a knee replacement or surgical operation, orthopaedic surgeons are required to monitor the flexion angle of the knee to track the patient's recovery progress. Currently, orthopaedic specialists utilize goniometers and rely on estimated passive observation to measure the joint's angle of flexion within its Range of Motion (ROM). However, goniometers have a high susceptibility to human judgment and errors when measuring the flexion angle and are invasive as the specialist needs physical contact with the patient to extract readings.
At DSAIL, we have developed a system based on mediapipe capable of processing real-time video input from a webcam and accurately determining the angle of flexion at the knee or arm joints. It does so by comparing the measurements obtained from a wearable flexion angle measurement device running on the same Jetson Nano.
Read more...