Project
Showcase
This application collects information on applicants’ academic background, professional experience, and program objectives to support the admissions review process and assess program fit.
When Complexity Fails: Why Simple Augmentation Outperforms CycleGAN for Pneumonia Detection
A rigorous failure analysis demonstrating that simpler data augmentation methods can outperform complex generative models in clinically sensitive medical imaging tasks.
Depth Completion Based on Stereo Disparity: A Comparison of Traditional Methods and Deep Learning
A comparative stereo vision pipeline showing how geometric methods and deep learning complement each other in robotic depth perception.
VQ-ViT Convolutional Fusion for Retinal OCT Anomaly Detection
An interpretable medical AI framework that balances diagnostic trust, accuracy, and real-world deployment efficiency.
Earthquake Rate and Average Magnitude Prediction Using XGBoost
Using machine learning and geospatial analysis to support earthquake preparedness and environmental risk mitigation.
Reconstructing 3D Spaces Using COLMAP: Understanding What Makes Reconstruction Succeed or Fail
A systematic exploration of how image quality, camera movement, and algorithm parameters shape 3D reconstruction accuracy.
Interactive Large-Scale 3D Reconstruction with Gaussian Splatting
An interactive 3D reconstruction system enabling scalable, immersive AR/VR experiences under real-world hardware constraints.
From Scratch Implementation of Visual SLAM with Graph Optimization
Building and evaluating a full visual SLAM pipeline to understand real-world robot localization and mapping limitations.
Smart Glasses for Brain Tumor Detection and Description from MRI Image
An AI-powered smart glasses system delivering real-time, hands-free brain tumor detection and interpretation during surgery.
Designing JP Morgan’s Innovation Engine for Scalable, Safe Transformation
A strategic blueprint that helps large banks innovate with fintech speed while remaining compliant and resilient.
Enhancing Biomedical Literature Retrieval with Semantic Textual Similarity
A domain-aware teacher–student model improving biomedical semantic similarity and clinical text retrieval.