My research focuses on making physical spaces machine-readable, converting 2D information into 3D models, detecting and processing industrial assets from point clouds, and applying AI to food safety certification for SMEs. Affiliated with ABIL Lab.
Best Paper Award, EUSPN 2025, Istanbul, Türkiye
🏆 Best Paper EUSPN 2025Open to collaborationsAvailable for mentorship
Six peer-reviewed publications, first author on three.
First author
Leveraging Machine Learning Techniques in Converting 2D Floorplans Images to 3D Models
Ibsen Giovanni BAZIE, Boaz N. Nzazi, Jirince K. Biaba, Tasho Tashev, Witesyavwirwa V. Kambale, Kyandoghere Kyamakya, Nathanael M. Kasoro, Selain K. Kasereka
With the growing demand for automation in fields such as architecture, real estate, and digital twin technologies, the ability to efficiently convert 2D floorplan images into accurate 3D structural models has become incr...
Tackling Japanese Traditional Floorplans: A Lightweight Segmentation Pipeline
Ibsen Giovanni BAZIE, et al.
iSCSi, Azores, 2026
Accepted
Enhancing Agricultural Supply Chain Traceability with Blockchain, Smart Contracts, and E-Labelling
Ibsen Giovanni BAZIE, Alidor M. Mbayandjambe, Darren Kevin T. Nguemdjom, Alain M. Kuyunsa, Hervek K. Kabengele, Rajesh Gupta, Tasho Tashev, Kyandoghere Kyamakya, Selain K. Kasereka
IEEE Big Data, Knowledge and Control Systems Engineering (BdKCSE), 2025
Africa's agricultural sector employs over 60% of the continent's population but faces challenges in product traceability and food security due to limited infrastructure and information asymmetries. This paper presents a ...
From IoT to AIoT: Evolving Agricultural Systems Through Intelligent Connectivity in Low-Income Countries
Selain K. Kasereka, Alidor M. Mbayandjambe, Ibsen G. Bazie, Heriol F. Zeufack, Okurwoth V. Ocama, Esteve Hassan, Kyandoghere Kyamakya, Tasho Tashev
Future Internet, MDPI, 2026
This paper explores the evolution from traditional Internet of Things (IoT) to Artificial Intelligence of Things (AIoT) in agricultural systems, with a specific focus on applications in low-income countries. The integrat...
Taxonomy Over Volume: Lessons from Building a 2D-to-3D Pipeline for Japanese Residential Floor Plans
Ibsen Giovanni BAZIE, et al.
Paper in preparation
In preparation
How you define the classes a vision system looks for matters more than model size or data volume. The paper works through this using the 13-class segmentation taxonomy built for the Akiya3D floorplan pipeline.
