Research Portfolio

Research Projects

Our projects span four research pillars, from AI-driven structural modeling to advanced materials and lifecycle management. Click any project to explore its architecture, methodology, and results.

8 projects

NeRF-UAVeL: Unified Attention-driven Volumetric Learning for Enhanced NeRF-based 3D Object Detection

AI-Driven Structural Performance3D Computer VisionActive

Unified attention-driven volumetric learning framework for state-of-the-art NeRF-based 3D object detection

Key Results

3D-FRONT dataset

3D-FRONT dataset

3D-FRONT dataset

3D-FRONT dataset

Novel Contributions

Multi-dimensional Volumetric Attention Pooling (MVAP)

Tri-Scale Asymmetric Convolutional Aggregation (TACA)

Dual-Domain Attention Fusion (DDAF)

NeRFPyTorch3D Object DetectionAttention MechanismsVolumetric Learning

View Full Details

GeoPAVE-AI: AI-Powered Assessment & Intelligent Maintenance of Georgia’s Concrete Pavements

PI: Dr. Tadesse Wakjira💰 $220,000 · Georgia DOT (GDOT)AI-Driven Structural PerformanceActive

Developing an AI-powered computer vision framework for automated concrete pavement condition assessment and intelligent maintenance planning.

Key Results

Georgia DOT sponsored

4-phase project

ASTM D6433 compliant

Decision-support prototype

Novel Contributions

End-to-End AI Pipeline

Decision-Support Tool

Georgia-Specific Dataset

Computer VisionDeep LearningAI/MLPCI / ASTM D6433Decision Support

View Full Details

PITCH: Physics-Informed Transformer for Hysteretic Response Prediction

PI: Dr. Tadesse WakjiraAI-Driven Structural PerformanceActive

A physics-informed transformer with token classification for hysteretic response prediction and uncertainty quantification of RC columns.

Key Results

Test median (n=38)

Novel Contributions

Soft-Decode Bridge

FiLM Conditioning

Native Uncertainty Quantification

TransformerGBMPhysics-Informed LossToken ClassificationPyTorch + CUDA

View Full Details

SA-EgoGS: Sharpness-Aware Dynamic Gaussian Splatting for High-Fidelity Egocentric Interaction

💰 IEEE Computational Intelligence SocietyAI-DrivenActive

Sharpness-aware dynamic 3D Gaussian Splatting for robust egocentric scene reconstruction with anti-aliasing and uncertainty-aware densification.

Key Results

3D Computer Vision

Sharpness-aware densification

Mip-style anti-aliasing

Novel Contributions

3D Computer Vision

Sharpness-Aware Densification (SHAD)

Scale-Aware Opacity Capping (SAOC)

3D Gaussian SplattingEgocentric VisionDynamic ReconstructionAnti-AliasingAlpha Compositing

View Full Details

ConceptCrack: Autonomous Crack Intelligence for Post-Disaster Structural Assessment

PI: Dr. Tadesse WakjiraAI-Driven Structural PerformanceActive

Autonomous AI pipeline fusing YOLO 26, SAM 3, and LLM for real-time crack detection and structural risk reporting.

Key Results

Hairline, flexural, shear, diagonal, wide, map cracking

Sub-mm pixel-precise masks

Risk scoring & remediation

Crack width compliance

Novel Contributions

Staged AI Fusion

Auto-Classification

LLM Structural Reasoning

YOLO 26SAM 3LLMOpenCVACI 224R

View Full Details

VulnCAST: Vulnerability & Capacity Assessment using Smart Transformers

PI: Dr. Tadesse WakjiraMulti-Hazard ResilienceActive

AI-powered seismic vulnerability and capacity assessment framework for reinforced concrete structures.

Key Results

Multi-hazard assessment

Regional risk analysis

Attention-based models

Damage probability curves

Novel Contributions

Transformer for Fragility

Multi-Scale Assessment

Interpretable Risk Models

TransformerXGBoostFragility AnalysisSHAPOpenSees

View Full Details

AI-StructVision: Teaching AI to Recognize Dangerous Structural Damage

PI: Dr. Tadesse Wakjira💰 $5,000 · KSU (Internal)AI-Driven Structural PerformanceActive

Can AI learn what ‘dangerous damage’ looks like in critical infrastructure?

Key Results

Bhatnagar, Boynton, Chet, Henock, Ruiz

Damage severity AI

Computer vision

Severity classifier

Novel Contributions

Severity-Aware Classification

Engineer-Guided AI Training

Ethical AI in Safety-Critical Systems

Computer VisionDeep LearningPythonImage ClassificationTransfer Learning

View Full Details

AI-Powered Comprehensive Framework for UHPC: From Material Optimization to Seismic Resilience of Bridges

PI: Dr. Tadesse WakjiraAI-Driven Structural PerformanceAdvanced MaterialsActive

A multi-scale AI framework spanning UHPC mix optimization, constitutive modeling, performance-based seismic design, and bridge resilience assessment.

Key Results

Top-tier journals

Multi-objective optimization

UHPC drift limit states

Most popular in journal

Novel Contributions

19-Objective UHPC Optimization

Generative AI for Constitutive Modeling

First UHPC Drift Limit States

Advanced ML ModelsCTGANSHAPMulti-Objective OptimizationOpenSees

View Full Details

Framework

Research Pillars

Collaborate With Us

Interested in our research? We welcome collaborations, sponsorships, and student inquiries.

Get in Touch Open Positions Google Scholar