PhD Student in Electrical and Computer Engineering at the University of Alabama.
I build deep learning pipelines for computer vision, generative AI, and time-series forecasting using PyTorch and TensorFlow.
| Project | What It Does | Stack |
|---|---|---|
| CLIP + LoRA Image Classification | End-to-end pipeline using CLIP embeddings with LoRA fine-tuning for efficient image classification | PyTorch, CLIP, LoRA |
| Conditional GAN Image Generation | Class-conditioned image synthesis on CIFAR-10 with gradient norm monitoring | TensorFlow, Keras |
| Food vs Non-Food Classification | CNN vs EfficientNetB0 transfer learning comparison on Food-5K | TensorFlow, Keras |
| Stock Price Forecasting | LSTM-based S&P 500 forecasting with SHAP feature selection | TensorFlow, SHAP |
| Handwritten Digit Classification | Neural network with experiments on initialization, optimizers, and K-Fold CV | TensorFlow, Keras |
ML & AI: Python, PyTorch, TensorFlow, Scikit-learn, NumPy, Pandas, CNN, GAN, LSTM, Transfer Learning, CLIP, LoRA Fine-Tuning, Vision-Language Models
Tools: Git, Jupyter, Google Colab, Linux, MATLAB
RF & EM: Ansys HFSS, Ansys Maxwell, Motor-CAD, Keysight ADS
Graduate Research Assistant at the University of Alabama β working on broadband antenna design, electromagnetic simulation, and permanent magnet optimization for motor systems using Ansys HFSS, Maxwell, and Motor-CAD.
Previously at EMMA Lab, Seoul National University of Science & Technology β designed RF transmission lines and connectors up to 67 GHz.
Ph.D. β Electrical & Computer Engineering, University of Alabama (in progress)
M.S. β Integrated IT Engineering, Seoul National University of Science & Technology
B.Sc. β Electronic & Telecommunication Engineering, IIUC, Bangladesh