Rehaan Chokshi completed Machine Learning for Automotive Safety

Built a perception-validation workflow that defined detections with IoU >= 0.5, computed precision/recall plus VRU false-negative rate, and sliced recall by lighting, visibility, and rarity to expose the highest-risk failure modes. The analysis correctly showed that night and occlusion sharply degrade VRU recall, and it prioritized those misses as the main safety concern.

• Leakage-Free Detection Definition
• Honest Safety Metric Evaluation
• Per-Condition Failure Analysis
• Risk-Based Triage