Welcome to DEPA Research Lab

Driving innovation at the intersection of technology and impact.

About DEPA Lab

The Data Engineering and Predictive Analytics Lab (DEPA Lab) at Morgan State University, led by Dr. Kofi Nyarko, is dedicated to unraveling the intricacies of complex systems and providing transformative insights. DEPA Lab focuses on applied research in Computer Vision, Machine Learning, and Artificial Intelligence Techniques.

Our Mission

We develop innovative machine learning models and algorithms for near real-time data collection, transformation, analysis, prediction, and visualization. DEPA Lab promotes inclusivity and innovation in data engineering and predictive analytics.

Our Vision

The Center for Equitable AI and Machine Learning Systems (CEAMLS) facilitates the development and deployment of socially responsible and equitable AI systems, ensuring they benefit everyone while educating the public about their impacts on health, prosperity, and happiness.

Core Research Areas

IoT Data Analytics

Analyze and visualize complex IoT networks for actionable insights.

Cybersecurity

Secure networks through advanced analytics and visualization techniques.

Machine Learning

Optimize algorithms for real-world decision support and autonomous systems.

Computer Vision

Automating traffic data analysis, pose estimation, and scene perception for diverse applications.

Ethical AI Framework

Advancing trustworthy and unbiased AI systems while addressing algorithmic bias.

Human-Computer Interaction

Creating interactive visualization tools for enhanced situational awareness.

Autonomous Systems

Developing robust algorithms for real-time navigation in challenging environments.

Innovative Projects and Research Highlights

At DEPA Research Lab, we are at the forefront of cutting-edge research, solving complex real-world challenges through interdisciplinary approaches. Explore some of our groundbreaking projects:

AI Assistive Comprehension Assessor

Developing an AI-powered tool to evaluate student comprehension by analyzing essays and generating quizzes in Canvas QTI/XML format, ensuring meaningful learning over AI-assisted responses.

Learn More

Benchmarking LLMs for AAVE & SAE

Assessing leading LLMs for their ability to generate dialect-specific text while maintaining semantic consistency, sentiment alignment, and reducing bias in AI-driven linguistic diversity.

Learn More

Ground Plane Segmentation for Wheelchair Mobility

Utilizing real-time segmentation techniques within the SAM framework to improve navigation for wheelchair users by accurately identifying traversable surfaces in diverse environments.

Learn More

AI/ML Bench Guard

A benchmarking framework for evaluating cloud-based and open-source ML services across tasks like object detection and facial recognition, promoting transparency and efficiency in AI service selection.

Learn More

Advanced Academic Support System

Integrating Generative AI into an academic advisory system to provide instant, AI-driven responses on course details, degree programs, and graduation requirements for student support.

Learn More

Queryable Computer Vision Pipeline

Developing a scalable system that allows data analysts to execute SQL-like queries on video content without deep computer vision expertise, bridging AI with real-world applications.

Learn More

AI Models in Action Recognition Video Analysis

The CLAIRE (Cross-Referencing Labels, Actions, and Interactions for Robust Explanations) project aims to integrate YOLO detections and LLM vision to analyze video frames.

Learn More

Crack Detection and Classification in Structural Materials

This research utilizes advanced machine learning techniques to detect cracks and structural anomalies in materials under stress.

Learn More

Optimizing Autonomous Systems with Edge Computing and CAD-Driven Robotics

This research utilizes edge computing and CAD modeling to develop real-time, on-device decision-making and innovative robotic solutions. By optimizing navigation algorithms and overcoming physical design limitations, it advances the capabilities of autonomous systems in complex, real-world environments.

Learn More

LLM Techniques for Multi-Object Tracking in Video Analysis

This research explores cutting-edge approaches to enhance the ability to track individuals across video frames accurately and consistently.

Learn More

Latest Events and News

Stay updated with the latest happenings, publications, and events at DEPA Research Lab.

DEPA Lab Award

DEPA Lab has been recognized for its groundbreaking research on equitable AI systems, fostering innovation in AI ethics and shaping policy frameworks for a fairer, more inclusive future.

Read More

Publications

DEPA Lab has made significant strides in advancing equitable AI systems, presenting groundbreaking research that addresses fairness, inclusivity, and transparency in artificial intelligence.

Read More

CEAMLS Symposium

Join DEPA at the CEAMLS Symposium as we delve into groundbreaking advancements in machine learning and AI. This annual event showcases DEPA's commitment to driving progress in AI and ML technologies.

Read More

Meet Our Team

Our diverse and talented team is committed to advancing research and innovation, solving real-world problems through technology.

Contact Us

Email: kofi.nyarko@morgan.edu

Address: Room 112 and 113 Schaefer Engineering Building, School of Engineering, 1700 E Cold Spring Ln, Baltimore, MD 21251