Hi,
DSAIL Mail is back! Did you miss us? Worry no more, welcome to our March and April newsletter! As we move further into the year, we are excited to share with you the latest updates, accomplishments, and upcoming events that have been taking place in our organisation :) |
Conferences and Workshops
TEMS AFRICA ICT EXPO 2024 |
On April 26th, Victoria Sitati from DSAIL gave a talk at the Sarit Centre Expo during the TEMS AFRICA ICT EXPO. Her presentation focused on the various ways in which technology can have a positive impact on the lives of individuals with disabilities, helping them recover and improve their overall well-being. The ICT Expo 2024 Africa focuses on converging the brightest minds, innovative technologies, and strategic partnerships in the field of Information and Communication Technology (ICT). |
|
|
Invited Guests
Andreas and Yimer |
On March 8th, we organised a technical session where we had the privilege of hosting Andreas Damianou, a Staff Research Scientist from Spotify. During his talk, he delved into the topics of Graph Learning and Graph Neural Networks, which proved to be extremely insightful. Furthermore, on March 25th, Yimer Asefa gave a talk on Weather Forecasting and Machine Learning, which was informative and well-received by the audience. We are glad to be part of such a vibrant community that brings together experts and enthusiasts from diverse fields. |
|
|
Visits
US Ambassador Visit |
During the visit of Ambassador Meg Whitman to Dedan Kimathi University of Technology, DSAIL members had the opportunity to share our work. We described our efforts developing solutions in the area of environmental conservation and healthcare. |
|
|
Research Progress
Low-Cost Non-Contact Forest Inventory: A Case Study of Kieni Forest in Kenya |
 |
We are thrilled to share that our research paper, entitled "Low-Cost Non-Contact Forest Inventory: A Case Study of Kieni Forest in Kenya," has been published in Challenges, an MDPI journal. As we all know, forests play a crucial role in providing us with food, fuel, medicine, and combating climate change. That's why having accurate data on forest resources is essential for making informed decisions. Unfortunately, traditional methods of gathering this data can be slow and laborious, while many non-contact methods in wide-spread use are typically expensive or less precise. With Kenya's ambitious goal of increasing forest cover to 30% by 2032, innovative technologies are needed to monitor the forests effectively. This is where our study comes in.
By utilising stereoscopic photogrammetry, we were able to extract accurate and timely data on tree heights and crown diameters in Kieni Forest, Kenya. This reliable method has the potential to support the national forest monitoring system and REDD+ implementation.
|
|
|
Research Progress
Lithium-ion Battery State of Charge Estimation |
The role of lithium-ion batteries in modern life is growing in areas such as grid energy storage and electric vehicles as the shift toward green energy accelerates. Monitoring the State of Charge (SoC) of lithium-ion batteries is essential for assessing their energy capacity. However, SoC cannot be directly measured; instead, it is estimated using parameters like voltage, current, and temperature.
The chemical and physical complexity of Lithium-ion Batteries increases the difficulty of making accurate predictions. In response to these challenges, Yuri Njathi, one of our researchers, is exploring machine learning methods to estimate SoC in battery packs. Machine learning offers a promising approach to handle multi-cell configurations and capture nonlinear relationships between input parameters and SoC. The goal is to design machine learning-based models that improve the accuracy and reliability of SoC predictions for complex battery systems
|
|
|
Competitions
IEEE Multi-Modal Video Reasoning and Analyzing Competition (MMVRAC) |
We are thrilled to announce that members of DSAIL took part in the IEEE ICME Grand Challenge MMVRAC and emerged as one of the top three teams in the Animal Pose Estimation Track.
We also wrote a workshop paper titled "AJA-pose: A Framework for Animal Pose Estimation based on VHR Network Architecture". Our Python open-source package, "AJA-pose," can be used to train and evaluate an animal pose estimation model using Vision Transformers and HRNet (VHR). Furthermore, we compared the results of our model with an earlier publication where the model was trained only on birds. Our findings indicate that our model outperforms when trained on all animals in the five vertebrates class. Being able to comprehend animals' poses holds great significance, ranging from identifying diseases in animals to detecting certain behaviours.
|
|
|
Upcoming Events
Data Science Africa |
The Data Science Africa event is right around the corner. The Summer School will take place from June 3rd to June 5th, while the Workshop will run on June 6th and 7th. This year's event promises to take your data science skills to the next level, with a range of exciting talks, workshops, and networking opportunities. If you're interested in learning more about the latest developments in data science and connecting with other industry professionals, DSA 2024 Nyeri is the place to be. Finally, a big congratulations goes out to all the accepted applicants - we can't wait to see you there! |
|
|
Spotlight
Congratulations to New Graduates |
Hats off to our very own Gabriel Kiarie (top left, MSc Telecommunication Engineering); Cedric Kiplimo (top right, MSc Telecommunication Engineering); Dishan Otieno (bottom left, BSc Mechatronic Engineering); and Emmanuel Ogal (bottom right, BSc Mathematics Modeling)! Here's to your incredible achievements and a bright future ahead! Wishing you all the best!🎉 |
|
|
Did a friend forward this to you? You can subscribe here
|
|
|
|
|
