Computer Vision • Deep Learning • Assistive Technology

Hi! I am Joana

A Computer Vision and Deep Learning Researcher

I am a Master in Computer Science student at the University of Wyoming working with Dr. Shivanand Venkanna Sheshappanavar in the Geometric Intelligence Research Lab. My work focuses on computer vision and deep learning for assistive technology, healthcare, and education.

Portrait of Joana Konadu Owusu

Recent News

Paper Accepted to WiCV at CVPR 2026 🎉

SAM3Count, have been accepted to the Women in Computer Vision (WiCV) Workshop at CVPR 2026 in Denver, Colorado, taking place June 3–7, 2026. I will be giving an oral presentation on this work, which focuses on open-vocabulary counting in images and videos under challenging conditions such as occlusion, fragmentation, and identity switches.

Projects

A few featured projects are highlighted below. More can be found on my GitHub page.

SAM3Count architecture diagram

SAM3Count

A zero-shot open-vocabulary counting framework for images and videos built on SAM3, with adaptive ROI-based refinement for dense scenes and stronger temporal consistency in videos.

Visit Repo
Self-driving car simulation on a road

Self Driving Car Simulation

A self-driving car simulation built with a feed-forward neural network, fitness-based search, and genetic algorithms to detect collisions and learn better paths.

Visit Live Site
XPT architecture diagram

XPT: Enhanced Point Transformer

An efficient and stable transformer architecture for 3D point cloud understanding that reduces computational cost and training time while maintaining strong accuracy.

Visit Repo
Mobile device

A Mobile App for the Visually Impaired

A mobile application that integrates computer vision and deep learning methods to support visually impaired users in real-world settings.

Visit Repo
Cats and dogs example

CNN for Cat and Dog Classification

A CNN-based classification project using audio spectrogram images derived from cat and dog sounds.

Visit Repo
Path visualization on a map

Optimal Path Finding for Emergency Response

Implemented AI search algorithms to estimate shortest paths for emergency flood response using Cheyenne highway maps.

Visit Repo

Work Experience

See more details on my CV or LinkedIn.

Graduate Research Assistant

Geometric Intelligence Research Lab - EECS

University of Wyoming

August 2024 - Present

  • Improve and develop VLM (Image, Video and 3D) and LLM models for assistive technologies, healthcare and autonomous driving.
  • Optimize and enhance efficiency of models and design hardware wearable IoT devices for persons with disabilities
  • Assisted in mentoring undergraduate students during REU NSF funded summer program

Software Development Specialist

Inclusive Tech Group

August 2020 - Present

A member of the product development team coordinating the design, development, testing, and implementation of assistive technology products for persons with disabilities. Learn more.

IT Coordinator

GES - Asuoso Community Senior High School

November 2021 - 2024

  • Administered and maintained school database systems and networks.
  • Performed network and system administration tasks on Windows Server and SAP software.
  • Improved system performance through proactive IT policy implementation.

Research & Administrative Assistant

Institute of Research, Innovation and Development - Kumasi Technical University

October 2020 - November 2021

Contributed to multidisciplinary research projects on food transportation and security in Ghana, perceptions of COVID-19 vaccines among staff and students, and the rise of street child begging in cities.

Education

University of Wyoming

Master of Science (MSc) Computer Science, 2024 - Present

GPA 4.0

University of Mines and Technology

Bachelor of Science (BSc) Computer Science & Engineering, 2016 - 2020

GPA 3.6

Courses

Analysis Of Algorithms Computer Animations Intro to Deep Learning Advances in Deep Learning Intro to LLMs Distributed Algorithms & Systems High Performance Computing

Skills

Languages: Python, C++, Java, JavaScript, SQL

ML / CV: PyTorch, TensorFlow, OpenCV, Transformers, Segmentation, Tracking

Systems: Linux, Git, Docker, Slurm, High Performance Computing

Focus Areas: Computer Vision, Deep Learning, Assistive Technology, Applied AI

Research & Publications

OVAC: Open Vocabulary Adaptive Counting in Images and Videos

Joana Konadu Owusu, Shivanand Venkanna Sheshappanavar • Under review for ICPR 2026

SAM3Count for Zero-Shot Open Vocabulary Counting in Images and Videos

Joana Konadu Owusu, Shivanand Venkanna Sheshappanavar • Oral • CVPR Workshop 2026

XPT: Enhanced Point Transformer V3

Research on improving the efficiency and stability of point transformer models for 3D point cloud understanding.

Nina: A Shopping Assistant Application for the Visually Impaired Using Computer Vision

Undergraduate thesis focused on assistive shopping support for visually impaired users.