Autonomous Vehicle Image Annotation for Object Detection
I participated in a large-scale data annotation project focused on training AI models for autonomous driving systems. My role involved labeling thousands of street-level images using bounding boxes and pixel-wise segmentation to identify and classify vehicles, pedestrians, traffic signs, road markings, and other key elements. Tasks required strict adherence to class hierarchy, annotation precision, and consistency across frames for tracking purposes. I maintained an average QA score of 96% and received consistent positive feedback from project leads for both speed and accuracy. Work was performed using Scale AI’s proprietary interface and CVAT for segmentation-based assignments. This project contributed directly to improving the visual recognition capabilities of next-generation self-driving vehicle models.