A curated collection of my open-source contributions, personal projects, and research initiatives.
MSc dissertation investigating privacy-preserving federated learning for Alzheimer's Disease classification using 3D MRI data from ADNI, introducing a novel Adaptive Local Differential Privacy mechanism.
Profile-aware RAG chatbot that uses small language models, retrieves vectorized resume/profile data, and serves responses locally via Ollama + Streamlit.
Benchmark study comparing UNet and DiT architectures for unconditional generation, with novel InfoNCE contrastive loss and SegFormer-based segmentation for attribute-conditioned face synthesis.
Benchmarking Modern LLMs on Ancient Buddhist Scripture Translation (Sanskrit/Pali to Vietnamese).
A comprehensive benchmark of 10 deep learning architectures for short-term solar forecasting in Ho Chi Minh City, prioritizing both accuracy and computational efficiency.
Novel OCR model for handwritten Chinese and Japanese text line recognition achieving SOTA accuracy, published in ICPR 2020.
Undergraduate Thesis. Built a Vietnamese dataset from MS-COCO and trained CNN+LSTM models.
Created a new Vietnam Traffic Signs Dataset and trained Faster RCNN model on 28 classes.