Projects
Brain Computer Interface System
This paper provides an in-depth review of methodologies for designing BCI-based applications,
focusing on the key challenges in brain signal acquisition, signal processing, and classification.
Various approaches are compared in terms of principles and performance, with recommendations
for optimal design choices to ensure functional and accurate application outcomes.
Object Detection Algorithms in Complex Scenes
C++, CUDA, Python, Matlab, TensorFlow/PyTorch, Keras, OpenCV, Numpy, Faster R-CNN, SQL, mAPUrban Parking Analytics: San Diego Case Study
This project analyzes San Diego's parking data to optimize urban mobility. Using the City's Open Data Portal, we examine parking meter transactions and locations to identify usage patterns, peak demands, and revenue trends. Through data analysis and visualization, we aim to enhance parking efficiency, reduce congestion, and improve resource allocation, contributing to a more accessible and efficiently managed urban environment.
Hate Speech Detector
Enhanced model balancing, expanded architectures, and improved evaluation metrics address dataset imbalance and boost detection accuracy on Twitter.
GNSS Accuracy: Kalman Filter
The project enhances smartphone GNSS accuracy to decimeter/
meter level in urban environments using Kalman Filter & Smoothing,
outperforming Weighted Linear Regression for precise positioning.
Robust Wheat Detector
EfficientDet/ Faster-RCNN with data augmentation techniques enhance
wheat head detection accuracy,
improving crop yield estimation across diverse environments.