AI Output Validator for Computer Vision Models
Validated the outputs of AI-generated object detection models in real-time mining operations. Provided structured feedback to improve model performance, focusing on detection accuracy and reducing false positives. Followed workflows similar to Reinforcement Learning from Human Feedback (RLHF) to guide iterative model refinement. • Evaluated and matched objects detected in computer vision outputs. • Used visual reasoning and structured assessment criteria for model validation. • Collaborated with automation teams to iterate improvements. • Enhanced operational safety by optimizing AI alert systems.