Ankita Singla completed Machine Learning for Automotive Safety

Built a complete perception validation dossier for a synthetic Aurora-7 detector: defined detection with an IoU >= 0.5 match rule, computed precision/recall and VRU false-negative rate, sliced recall by night/occlusion/rarity, and ranked the highest-risk failure modes. Finished with an internally consistent GO-WITH-CONDITIONS verdict and a SOTIF safety case that tied the release decision to the student’s own metrics and the real Uber Tempe crash.

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