PhD Student · Computer Science · Georgia State University
ML researcher working on medical imaging and human trajectory prediction.
Research
PhDOriginal work in medical imaging and human motion prediction.
Brain Tumor Inpainting
Deep learning system for synthesizing healthy brain tissue from tumor-affected MRI scans. Given only a corrupted scan, the model infers what the underlying anatomy should look like — supporting non-invasive assessment and surgical planning without requiring paired healthy images.
Details
- Multi-modal MRI input (T1, T1ce, T2, FLAIR modalities)
- Trained and evaluated on BraTS 2023 Adult Glioma dataset
- Evaluated using masked PSNR — the BraTS 2023 winner protocol
Masked PSNR
23.41 dB
SSIM
0.93
LPIPS
0.08
Dataset
BraTS 2023
Multi-Agent Trajectory Prediction
Deep learning model for predicting multiple plausible future trajectories for agents in crowded scenes, conditioned on observed motion history and neighboring agent behavior. Evaluated on standard autonomous driving and pedestrian simulation benchmarks.
Details
- Benchmarked on ETH / UCY pedestrian prediction datasets
- Probabilistic output — K diverse future trajectories per agent
- Supports up to 20 neighboring agents per scene
Metric
minADE / minFDE
Samples
K = 20
Benchmark
ETH / UCY
Horizon
8 obs → 12 pred
Projects
Applied engineering and data science across NLP, computer vision, and systems.
xFPL
Dec 2025Fantasy Football · Full-Stack App
Solo-built AI-powered Fantasy Premier League management system with squad building (£100M budget), captain system, head-to-head leagues, and live gameweek simulation. AI transfer recommender scores candidates by recent form, fixture difficulty rating (FDR), and points-per-million value.
2nd Best Project · Graduate Category · DB Systems, Fall 2025
LingoScape
Nov 2024Multi-Modal Translation Platform
Full-stack multilingual platform supporting 15 languages with real-time speech-to-text, text translation, text-to-speech, and video conferencing. Built with a 5-person team — placed 2nd among 28 teams at the GSU CS demo competition.
2nd / 28 teams · <500ms latency
Brain Tumor Detection
Jul 2024Object Detection · RetinaNet
Trained RetinaNet (ResNet-50 + FPN) on 1,229 brain MRI scans using Detectron2. Achieved state-of-the-art detection accuracy across multiple IoU thresholds on Google Colab Pro with A100 GPUs.
94% mAP @ IoU 0.5 · 91% @ IoU 0.75
US Airline Sentiment Analysis
Jul 2023NLP · Twitter Feedback
Sentiment classifier for airline customer feedback using logistic regression on 14,640 tweets. Expanded feature dimensionality from 8.5K to 61K via TF-IDF and n-gram preprocessing.
78.9% accuracy · 84% precision on negatives
Student Hostel Price Prediction
2022Undergraduate Research · KNUST
First-ever ML study of the KNUST student hostel market. Manually collected 500 responses from 70 hostels and trained regression and neural network models to predict room prices from location, amenities, and proximity features.
R² > 0.75 · 77%+ accuracy across all models
Experience
Three years of teaching and mentoring, staying in close contact with 6 students.
Graduate Teaching Assistant
Theory Foundations of Computer Science
Georgia State University
- Evaluated assignments for over 70 students, providing personalized feedback and guidance.
- Led weekly lab sessions to reinforce core concepts in computability, complexity, and formal languages.
Graduate Teaching Assistant
Foundations of Data Science
Georgia State University
- Evaluated data science assignments for over 55 students with detailed, individualized feedback.
- Led weekly office hours mentoring students in Python and data visualization — average homework scores rose from 88.9% to 96.7%.
Graduate Lab Assistant
Elementary Statistics
Georgia State University
- Visualized statistical concepts to help over 200 students build intuition for key ideas.
- Documented common challenges and effective teaching strategies, contributing to an 8% improvement in student performance.
Awards & Skills
Recognition
2nd Place — CS Demo Competition
Led a 5-person team to build LingoScape, a multi-modal language translation app, placing 2nd among 28 teams.
Top Graduate Lab Assistant
Selected as top performer among 15 Graduate Lab Assistants for outstanding student mentorship at Commons Math Lab.
Willey M. Suttle Math Award
Recipient of the departmental award for academic excellence in mathematics.
Technical Skills
Languages & Tools
Machine Learning
Data & Visualization
About
I'm a Computer Science PhD student at Georgia State University, where I also hold an MS in Mathematics & Computer Science (GPA 3.81, Scientific Computing track). My current research focuses on two problems: synthesizing healthy brain tissue from tumor-affected MRI scans, and predicting plausible future trajectories for agents in crowded scenes. Both sit at the intersection of deep learning and real-world perception, and both are works in progress toward publication.
Before the PhD, I developed a strong foundation in statistical modeling, NLP, and computer vision through graduate coursework and independent projects — from RetinaNet-based tumor detection (94% mAP) to logistic regression sentiment analysis on 14K+ tweets to the first ML study of student hostel pricing at KNUST, Ghana. That last project, which I led as an undergrad, taught me that the hardest part of applied ML is often the data — we manually collected 500 survey responses across 70 hostels.
I've spent three years teaching and mentoring — staying in close contact with 6 students across statistics, data science, and CS theory — and that work has shaped how I think about research communication. I write code and papers with the same goal: make the idea clear enough that someone else can build on it. I'm looking for research collaborations, internships, and roles in ML engineering or applied AI where that combination of rigor and clarity is valued.
Program
PhD · Computer Science
School
Georgia State
Expected
May 2030
GPA
3.81 (MS)
Reading
Reading
Read
Contact
Open to research collaborations, internships, and full-time ML/AI roles.