Fields of interest: Computer Vision, Mechatronics and Robotics, Multibody Dynamics Simulation
- Languages: Python, JavaScript, C++
- ML/DL: OpenCV, PyTorch, Tensorflow, Lightning, Keras, Scikit-learn
- DevOps: Linux, Git, Docker
- Robotics: Webots simulator, Exudyn dynamics simulator
- CAD: Solid Edge, Inventor, Alibre
-
Automation of TV remote controls to assist older adults with button visibility and navigation.
-
A document scanner application that converts images of rectangular documents into a top-down view with classical computer vision, tested in a robotics simulator.
-
30K cards identification system
Card identification system that matches a query card to a 30K card database using fine-tuned deep feature embeddings and similarity search.
-
Semantic segmentation on images taken from drones to differentiate between 12 classes.
-
Object detection to localize license plates of cars.
-
Image classification of 13 different Kenyan food types.