Obed Mogaka, Doctoral student

I am a Doctoral student at Mälardalen University, within the School of Innovation, Design, and Engineering, specifically in the Division of Intelligent Future Technologies (IDT) under the Academy for Innovation, Design, and Technology. I am supervised by Professor Masoud Daneshtalab.

My research passion lies in leveraging embedded systems to address human-centered challenges. I am particularly interested in the efficient inference of AI on resource-constrained embedded devices, with a current focus on accelerating deep learning using Field Programmable Gate Arrays (FPGA).

I hold an MSc in Electronics and Communication Engineering from the Egypt-Japan University of Science and Technology (2024) and a BSc in Electrical and Electronics Engineering from Jomo Kenyatta University of Science and Technology, Kenya (2019).

My research focuses on the efficient inference of deep learning algorithms on resource-constrained embedded devices. I am currently developing methods to design energy-efficient deep learning models with low energy consumption and memory usage, enabling the rapid implementation of complex models on various edge devices while adhering to specific hardware constraints.

My specific areas of interest include: 

• Heterogeneous Computing: Hardware/Software Co-design
• FPGA-based Accelerator Design
• Efficient Deep Learning and TinyML
• Embedded Systems