Vishesh Satarkar completed Machine Learning for Automotive Safety

Built a complete perception-validation dossier for a synthetic Aurora-7 detector: defined detection by IoU≥0.5, scored precision/recall and VRU false-negative rate, and then broke VRU recall down by lighting, visibility, and class rarity to surface the highest-risk slices. The final release verdict was evidence-based and internally consistent, with SOTIF-style mitigations and a grounded comparison to the Uber Tempe crash.

• Leakage-Free Metric Definition
• Safety-Critical Slice Analysis
• ISO 21448 SOTIF Reasoning
• Evidence-Based Release Decisioning