Each summer, IMPACT hosts a number of interns. This year represented a record number, and over the next couple of weeks we will be telling you the stories of their experience. This first post spotlights Shreya Shrestha and Deepak Menon. Both interns — Shreya, a student at Vanderbilt University, and Deepak, a high school senior — worked with IMPACT team member Slesa Adhikari on a project that created a Python package for the ImageLabeler tool.
The ImageLabeler Python package allows users to interface with the ImageLabeler tool as they label datasets for use in machine learning training. It provides users the ability to work with images, GeoTIFFs, and shapefiles related to Earth science events more easily by removing the need to interact with the graphical user interface (GUI). Datasets for machine learning projects are comparatively quite large, so it is not always practical for a user to interact solely with the GUI, especially as it can become a tedious process. This package provides a secondary interface that eases the process of preparing and interacting with datasets.
IMPACT’s ImageLabeler tool was covered in depth in a previous blog post. Though developed to facilitate machine learning in the Earth sciences, the ImageLabeler product is broadly applicable across domains and one doesn’t have to be from an Earth science background to be able to use it. As Deepak noted, the Python package will help research scientists utilize ImageLabeler capabilities with more ease.
