Generate robust training dataset for alignment from existing 3D meshes

Test network models in production-level pipeline using C++, OpenCV, and AWS Batch Job

Reorganize codebase and set up testing in Jenkins for quality control on research projects

Tracked camera pose and trajectory with DSTN neural network for key point detection

Removed background in film with UNet network architecture for alpha matting and segmentation

Improved training results by applying image augmentations in OpenGL (e.g. noise and color jitter)

Developed online training using gRPC to send updated ground truth data from a GUI front end

Developed real-time video processing pipeline in an interactive front end with ImGUI and C++

Created web dashboard to show test information and related tests for workflow management

Tested and improved regression unit tests for DB2/zOS utilities

Implemented ASCII to EBCDIC conversion script to assist in creating new unit tests