Home

Md Raihan Khan

I am a Lecturer in the Department of Electrical and Electronic Engineering at North Western University, Bangladesh, and an B.Sc. graduate in EEE at the Khulna University of Engineering and Technology (KUET), advised by Prof. Mohiuddin Ahmad.

My research develops efficient, geometry-aware deep learning with a focus on:

• One- and few-step generative modeling — closed-form flow and optimal-transport formulations that collapse iterative sampling into a single network evaluation
• Image restoration and exposure correction under real-world degradation, formulated through principled optimal transport
• Geometry-aware learning on Riemannian / SPD manifolds — equivariant architectures, geodesic operators, and numerically stable matrix-function approximation
• Structured representation learning for neural and biomedical signals, including EEG-based affective and cognitive state decoding

I am applying to PhD programs for the 2027 cycle in Machine Learning / Electrical and Computer Science, and I am actively looking for advisors working on generative modeling, optimal transport, and geometric deep learning.

Download CV (PDF) Publications Google Scholar

Selected recent work

• Orthogonal Polynomial Approximation for Matrix-Log Normalization in Global Covariance Pooling — BMVC 2026 (accepted). GPU-friendly polynomial approximations that move a previously CPU-bound operation onto the GPU.
• AutoLumNet: Monotone Optimal Transport for Single-Shot Exposure Correction — IEEE TPAMI (under review, 2026).
• Equivariant Geodesic Networks: End-to-End Classification on the SPD Manifold — AAAI 2026 (under review).
• DAFNet: A Dual-Path Attention Fusion Network for EEG Emotion Recognition — Array (Elsevier), 2025.

See the Publications page for the full list.

Contact

• Email: kraihan918@gmail.com
• CV · Publications · Portfolio · Teaching