Blessing Agyei Kyem

I am a second year PhD student at the Sustainable Mobility and Advanced Research in Transportation (SMART) Lab at North Dakota State University, where I work on Multi-modal AI in Computer Vision, Agentic AI and their applications in Transportation and Pavement Engineering. My PhD advisor is Armstrong Aboah.

Email  /  GitHub  /  Google Scholar  /  ResearchGate  /  Blog  /  X  /  LinkedIn  / 

profile photo

News



[10/05/2025]         Our paper titled A Memory-Augmented Transformer with Multi-Scale Spatio-Temporal Modelling for Network-Wide Traffic Flow Prediction has been accepted for presentation at TRB 2026

[10/03/2025]         Our paper Self-Supervised Multi-Scale Transformer with Attention-Guided Fusion For Efficient Crack Detection has been accepted for publication at Automation in Construction

[09/30/2025]         Demographics-Informed Neural Network for Multi-Modal Spatiotemporal Forecasting of Urban Growth and Travel Patterns Using Satellite Imagery has been accepted for presentation at TRB 2026

[09/29/2025]         Our paper Demographics-Informed Neural Network for Multi-Modal Spatiotemporal Forecasting of Urban Growth and Travel Patterns Using Satellite Imagery has been accepted In NeurIPS 2025 Workshop on UrbanAI!

[09/28/2025]         Our paper Self-Supervised Multi-Scale Transformer with Attention-Guided Fusion For Efficient Crack Detection has been accepted for presentation at the TRB 2026 Conference

[09/28/2025]         Context-CrackNet: A Context-Aware Framework for Precise Segmentation of Tiny Cracks in Pavement Images has been accepted for presentation at the TRB 2026 Conference

[09/20/2025]         Our paper An Enhanced Lightweight Model for Real-time Pavement Condition Index Prediction has been accepted for presentation at the TRB 2026 Conference

[07/12/2025]         Three of our research papers have been accepted at ICCV 2025!

[04/29/2025]         Our research paper: Saam-Reflectnet: Sign-Aware Attention-Based Multitasking Framework for Integrated Traffic Sign Detection and Retroreflectivity Estimation has been accepted for publication at Expert Systems with Applications

[04/29/2025]         Our research paper: Context-CrackNet: A Context-Aware Framework for Precise Segmentation of Tiny Cracks in Pavement Images has been accepted for publication at Construction and Building Materials

[02/26/2025]         Our paper A Big Data Approach to Pavement Distress Detection has been accepted for oral presentation at the International Conference on Big Data Analytics 2025

[02/26/2025]         Our paper Integrating Travel Behavior Forecasting and Generative Modeling for Predicting Future Urban Mobility and Spatial Transformations has been accepted for oral presentation at the International Conference on Big Data Analytics 2025

[12/19/2024]         Our paper titled Weather-Adaptive Synthetic Data Generation for Enhanced Power Line Inspection Using StarGAN has been accepted for publication at IEEE Access

[09/25/2024]         Our research paper titled A novel methodological framework for assessing traffic sign retroreflectivity using Lidar data has been accepted for presentation at the TRB 2025 Conference

[09/22/2024]         Our paper PaveCap: The First Multimodal Framework for Comprehensive Pavement Condition Assessment with Dense Captioning and PCI Estimation has been accepted for presentation at the TRB 2025 Annual Conference Meeting

Research

I'm interested in deep learning, computer vision, LLMs, multimodal LLMs, and agentic ai.

project image

Task-Specific Dual-Model Framework for Comprehensive Traffic Safety Video Description and Analysis



Blessing Agyei Kyem, Neema Jakisa Owor, Andrews Danyo, Joshua Kofi Asamoah, Eugene Denteh, Tanner Muturi, Anthony Dontoh, Yaw Adu-Gyamfi, Armstrong Aboah
ICCV, 2025
paper / arXiV / slides / post /

We propose a dual-model pipeline for traffic-scene video understanding that narrates safety-relevant events and answers questions about them using multimodal LLMs.

project image

Prompt-Guided Spatial Understanding with RGB-D Transformers for Fine-Grained Object Relation Reasoning



Tanner Muturi, Blessing Agyei Kyem, Joshua Kofi Asamoah, Neema Jakisa Owor, Richard Dyzinela, Andrews Danyo, Yaw Adu-Gyamfi, Armstrong Aboah
ICCV, 2025
paper / arXiV / post

We present a prompt-guided RGB-D multimodal LLM that performs fine-grained spatial reasoning in warehouse environments. It integrates object bounding boxes and depth cues into textual prompts to improve relation reasoning, distance estimation, and spatial grounding.

project image

A Unified Detection Pipeline for Robust Object Detection in Fisheye-Based Traffic Surveillance



Neema Jakisa Owor, Joshua Kofi Asamoah, Tanner Wambui Muturi, Anneliese Jakisa Owor, Blessing Agyei Kyem, Andrews Danyo, Yaw Adu-Gyamfi, Armstrong Aboah
ICCV, 2025
paper / arXiV / post

We propose a unified fisheye objection detection pipeline that combines fisheye-aware preprocessing, super-resolution/slicing postprocessing, and an ensemble of modern detectors to robustly detect small and distorted traffic objects.

project image

Self-supervised multi-scale transformer with Attention-Guided Fusion for efficient crack detection



Blessing Agyei Kyem, Joshua Kofi Asamoah, Eugene Denteh, Andrews Danyo, Armstrong Aboah
Automation in Construction, 2025
paper / arXiV / code / bibtex

We present Crack-Segmenter, a self-supervised multi-scale transformer for pixel-level crack segmentation.

project image

Saam-Reflectnet: Sign-Aware Attention-Based Multitasking Framework for Integrated Traffic Sign Detection and Retroreflectivity Estimation



Joshua Kofi Asamoah, Blessing Agyei Kyem, Nathan-David Obeng-Amoako, Armstrong Aboah
Expert Systems with Applications, 2025
paper / SSRN / bibtex

This research involves designing a multi-tasking framework, called Saam-ReflectNet, that combines traffic sign detection, classification, and retroreflectivity estimation into a single workflow.

project image

Context-CrackNet: A Context-Aware Framework for Precise Segmentation of Tiny Cracks in Pavement images



Blessing Agyei Kyem, Joshua Kofi Asamoah, Armstrong Aboah
Construction and Building Materials, 2025
paper / arXiv / bibtex

This study introduces Context-CrackNet, a novel architecture combining Region-Focused Enhancement Module (RFEM) and Context-Aware Global Module (CAGM) to improve crack segmentation.

project image

Weather-Adaptive Synthetic Data Generation for Enhanced Power Line Inspection Using StarGAN



Blessing Agyei Kyem, Joshua Kofi Asamoah, Armstrong Aboah, Ying Huang
IEEE Access, 2024
paper / bibtex

This research introduces a framework combining heuristic image processing and StarGAN to generate synthetic power line inspection images under adverse weather conditions like fog, rain, and nighttime.

project image

PaveCap: The First Multimodal Framework for Comprehensive Pavement Condition Assessment with Dense Captioning and PCI Estimation



Blessing Agyei Kyem, Eugene Kofi Okrah Denteh, Joshua Kofi Asamoah, Armstrong Aboah
arXiv, 2024
arxiv / code / poster / website /

We show how to provide dense caption describing the condition of a pavement image. The textual description captures the types of pavement distress present with their severity, the absent pavement distresses and the pavement condition index (PCI)

project image

Advancing Pavement Distress Detection in Developing Countries: A Novel Deep Learning Approach with Locally-Collected Datasets



Blessing Agyei Kyem, Eugene Kofi Okrah Denteh, Joshua Kofi Asamoah, Kenneth Adomako Tutu, Armstrong Aboah
arXiv, 2024
arxiv / website /

This research presents a deep learning approach combining YOLOv5 with Convolutional Block Attention Module (CBAM) to detect and classify pavement distress types in developing countries.


Design and source code from Jon Barron's website