Project description. This is a skill-building project focused on mastering pixel-perfect semantic segmentation. The objective was to achieve high-precision results for potential machine learning applications. I developed a workflow involving: Annotation: Using LabelMe to create precise polygons. Automation: Writing and testing Python scripts to fix polygon overlapping and pixel gaps in the JSON datasets. Quality Control: Achieving a 99.6% IoU score through rigorous audit processes.
- Data Annotator