Programming Languages: Python, C++, JavaScript, SQL, R
Frameworks: PyTorch, TensorFlow, Scikit-learn, NumPy, Pandas, OpenCV, Torchvision, Transformers
Technologies: AWS (S3, ECS), Docker, ONNX, TensorRT, CUDA, SLURM, Conda, Linux, Spark
Visit my Github to view more of my projects.
AI Paper Similarity Search App
A web app to help discover similar arXiv papers using fine-tuned Sentence-BERT embeddings trained on 2 million papers. Fully AI-powered search engine with IndexedDB for fast, serverless searches. Built using Transformers.js for efficient academic paper recommendations.
Digits Object Detection App
An interactive web app for real-time digit detection using OpenCv.js and a custom trained YOLOv8 model. Fully optimized with ONNX for seamless and efficient digit recognition.
Contrastive Pre-Training and Multiple Instance Learning for Predicting Tumor Microsatellite Instability
DOI: 10.1109/EMBC53108.2024.10782037
Ronald Nap, Mohammed Aburidi, Roummel Marcia
Engineering in Medicine and Biology Society (EMBC) 2024
Synthetic Data Generation for Deep Learning Model Enhancement
A proof-of-concept research project leveraging Generative Adversarial Networks to generate synthetic data for improving classification performance. By iteratively refining images with a CNN classifier, the project enhances the accuracy and robustness of emotion recognition. The synthetic data supplements the limited dataset, effectively boosting model performance.