Anuj Kale completed Machine Learning for Automotive Safety

Built a perception validation dossier end to end for a synthetic Aurora-7 detector: defined detection with an IoU >= 0.5 rule, computed precision/recall and the VRU false-negative rate from scored detections, and sliced recall by lighting, visibility, and class to expose the night/occluded failure modes. Closed with ISO 21448-style SOTIF chains, a daylight-focused GO-WITH-CONDITIONS recommendation, and a grounded comparison to the Uber Tempe crash.

• Leakage-Free Detection Definition
• Safety-Critical Metric Selection
• Per-Condition Failure Triage
• ISO 21448 SOTIF Reasoning