01

About

Yunusa Haruna
📍 Oulu, Finland
GitHub

Yunusa is a postdoctoral researcher at the Center for Machine Vision and Signal Analysis (CMVS), University of Oulu, Finland, and a visiting researcher with the Machine Learning Group at UiT The Arctic University of Norway. His work at CMVS focuses on developing efficient alternatives to Transformer models using physically inspired dynamical systems, particularly state-space models to improve long-range dependency learning in visual data, with applications to multimodal medical imaging.

His notable work includes vGamba, with publications in top-tier venues such as TMLR, ICCV, and COLING, alongside several SCI-indexed journals and conferences. He obtained his PhD in Pattern Recognition and Intelligent Systems from Beihang University, Beijing, China. During his doctoral studies, he interned at X-Lab (Tsinghua University) in popular robotics, Pegasus AI, and China Mobile International. He is also the founder of NewraLab, Suzhou, China; an AI research and development lab focused on lightweight, edge-ready architectures for emerging regions and low-resource environments.

He holds an MSc in Computer Information Systems from Yakın Doğu Üniversitesi, Nicosia, Cyprus and a BSc in Computer Science from Bayero University, Kano, Nigeria. He has mentored PhD and master's students and contributed to research supervision in computer vision and deep learning. Outside research, it's coffee, cooking, football and boxing.

current_focus.sh
$positionPostdoc · CMVS, University of Oulu, Finland · Visiting, UiT Norway
$researchState-Space Models · Efficient Vision · Medical Imaging
$notablevGamba · iiANET
$statusopen to collaboration
02

News

May 2025

🎯 Excited! Joining Qing Liu's research group at CMVS, University of Oulu as a postdoctoral researcher in medical imaging.

Jan. 2026

🎉 Thrilled! iiANET accepted at TMLR — Transactions on Machine Learning Research. Years of work, finally in print.

Mar. 2025

🚀 Excited! Selected for Google's TPU Research Cloud (TRC) Program, gaining access to TPU infrastructure to accelerate research in VLMs.

Feb. 2025

🙌 Excited! NewraLab selected for the NVIDIA Inception Program, joining NVIDIA's global startup ecosystem for AI companies.

Mar. 2025

Our work on the synergies of hybrid Vision Transformers and CNNs was published in Engineering Applications of Artificial Intelligence. [link →]

Dec. 2024

MambaForGCN accepted at COLING 2025 — enhancing long-range dependency with SSM and Kolmogorov-Arnold Networks for ABSA. [link →]

Oct. 2024

KonvLiNA presented at ICMV 2024, Edinburgh — integrating KAN with Linear Nyström Attention for crop field detection. [link →]

Dec. 2023

Presented iiANET at ACCV Workshop — Computer Vision for Developing Countries, Hanoi, Vietnam. [slides →]

03

Publications

Journal
Exploring the Synergies of Hybrid CNNs and ViTs Architectures for Computer Vision: A Survey
Conf.
MambaForGCN: Enhancing Long-Range Dependency with State Space Model and Kolmogorov-Arnold Networks for Aspect-Based Sentiment Analysis
Journal
Multi-Scale Dual-Stream Visual Feature Extraction and Graph Reasoning for Visual Question Answering
Journal
SaRPFF: Self-Attention with Register-Based Pyramid Feature Fusion for Enhanced Rice Leaf Disease Detection
$ visitor_map --live