HITL (Human-in-the-Loop) Expert — Human-in-the-Loop at scale for AV computer vision model validation and tracking refinement
Provided Human-in-the-Loop (HITL) validation and refinement for computer vision models used in autonomous vehicle (AV) navigation. Performed image-to-video tracking work by identifying dynamic objects in real-world environments to improve model safety and accuracy. Ensured labeled/tracked outputs supported reliable self-driving algorithm performance in operational settings. • Validated computer vision model predictions in a HITL workflow • Refined tracking quality for dynamic object identification • Used CVAT to manage complex image-to-video tracking tasks • Focused on safety and accuracy for AV navigation scenarios