Focused on developing computer vision solutions using classical and deep learning-based methods.- Developing and training custom object detection models. Debugging models and designing hyperparameters in accordance with the relevant performance metrics.- Planning collection and preparation of datasets for object detection tasks. Designing and building custom tools for dataset collection & preparation and metrics logging.- Evaluating the deep learning-based approaches to the visual object tracking problem wrt. to their ease of implementation, performance.- Building end-to-end image stitching pipelines for a panoramic IR camera system, including preprocessing, feature matching, K-R-T estimation, bundle adjustment, seam estimation.- Inertial- and GPS-aided SLAM and VIO systems for use in urban environments.