Curriculum Vitae — Tahir Hasan

Education

MSc Computer Vision, Robotics and Machine Learning Sep 2025 – Sep 2026 (expected) University of Surrey — Guildford, UK

Semester 1 average: 90.5%.
Modules: Advanced Topics in Computer Vision & Deep Learning, Computer Vision & Pattern Recognition, Robotics, Medical Robotics, Applied Machine Learning, Speech & Audio Processing & Recognition, Artificial Intelligence of Things.

BSc Mechatronics Engineering Feb 2019 – May 2024 Rajshahi University of Engineering & Technology (RUET) — Rajshahi, Bangladesh

CGPA 3.54 / 4.00.
Thesis: Implementation of Bangla Sign Language (BdSL) Recognition in Continuous Real Time — now a paper under review.

Research Assistant Feb 2024 – present AMIR Lab — Advanced Machine Intelligence Research Lab — Rajshahi, Bangladesh (remote)

Developing a hierarchical coarse-to-fine framework for gastrointestinal image analysis: a router model resolves anatomical origin (upper vs. lower GI) before location-specific expert models classify findings — improving diagnostic accuracy by over 12%.
Extensive experiments in PyTorch with customised CNN backbones (ResNet/EfficientNet).
First-authored an explainable two-stage plant-disease identification pipeline (YOLO + ECA-NFNet), published at IEEE ICCIT 2025.

Undergraduate Researcher (thesis) Jan 2023 – May 2024 Dept. of Mechatronics Engineering, RUET — advisor: Sumaya Ishrat Moyeen — Rajshahi, Bangladesh

Built a real-time continuous BdSL recognition system: MediaPipe Holistic landmarks feeding LSTMs, with an attention mechanism that improves sentence-formation accuracy.
Curated a novel video dataset — 9,010 clips across 530 word classes — addressing the scarcity of BdSL resources.

Lecturer Feb 2025 – Aug 2025 Varendra University, Dept. of Computer Science and Engineering — Rajshahi, Bangladesh

Taught undergraduate Computer Algorithms and System Analysis and Design; designed the curriculum and laboratory modules for both.
Mentored student projects in machine learning and applied AI.

Industrial Trainee Mar 2022 – Apr 2022 Monno Ceramic Industries Ltd. — Dhaka, Bangladesh

Analysed automated production systems, including PLC-controlled kilns and power-generation units.
Assisted the engineering team with maintenance and fault detection on robotic glazing arms.

Hierarchical Coarse-to-Fine GI Image Analysis — A router network sends endoscopy images to specialist classifiers — +12% diagnostic accuracy over a flat single-stage baseline.
Continuous Real-Time BdSL Recognition — Real-time recognition of continuous Bangla Sign Language, built on a purpose-collected 9,010-video, 530-class dataset.
Plant Disease 2-Stage XAI Pipeline — YOLO detection feeding an ECA-NFNet classifier with explainability overlays — published at IEEE ICCIT 2025.
Full list, including earlier robotics and embedded work: projects page.

Languages & frameworks: Python (PyTorch, TensorFlow, OpenCV, Pandas) · C/C++ · MATLAB · SQL · Java
Vision specialisms: Detection (YOLO) · CNN backbones (NFNet, ResNet, EfficientNet) · MediaPipe landmark pipelines · LSTM/attention temporal models · CAM-based explainability
Tools & design: Git · Docker · ROS 2 · SolidWorks · AutoCAD · Proteus · Unity
English: IELTS 8.5 overall (R 9.0 · L 9.0 · S 8.0 · W 7.5)

University Innovation Hub Program (UIHP) pre-seed grant — ICT Division, Govt. of Bangladesh, 2024: automated monitoring systems for a hydroponics startup.
Top Innovator, Banglalink Ennovator 7.0 (2023) — top finalist for AI-driven social-development solutions.
IELTS 8.5 overall.