প্রজেক্ট
Automated MCQ Generation with Options and Answering based on Text and Images using NLP
Final year project generating Bangla MCQs from text/images using fine-tuned LLMs, with a Streamlit quiz UI.
Md Abdullah Al Maruf
B.Sc. · Computer Science and Engineering (CSE)
Daffodil International University
প্রজেক্ট
Final year project generating Bangla MCQs from text/images using fine-tuned LLMs, with a Streamlit quiz UI.
প্রজেক্ট
Lets users link bKash, Nagad, Rocket, and Upay accounts and transfer funds between them from one app.
প্রজেক্ট
A real-time 2D boat survival game rendered from raw OpenGL primitives — no engine, no sprites.
প্রজেক্ট
Classifies vegetable photos into 15 categories with 99.97% validation accuracy, deployed via Streamlit.
A computer vision project that fine-tunes a ResNet18 model to classify vegetable images into 15 classes, trained on roughly 18,000 images and reaching 99.97% validation accuracy. I built a Streamlit front end where users upload a photo and get an instant prediction, with the trained model auto-downloaded from Google Drive at runtime so the repo stays lightweight. This has been a centerpiece project when applying for AI/ML Engineer (Computer Vision) roles.