Data Labeler for Accident Detection Project
The project involved training a YOLO model to detect accidents in real-time video footage. I labeled a custom dataset with accident and non-accident scenarios to train and validate the model. The labeling process focused on identifying vehicles and analyzing their spatial relationships. • Prepared and labeled video clips for machine learning use. • Ensured high-quality annotations distinguishing between accident and non-accident events. • Used a consistent schema to support model training and evaluation. • Contributed to overall system performance improvements.