Sharvari Nene completed Machine Learning for Automotive Safety

Built a complete perception-validation dossier for a fictional Aurora-7 detector: defined detection correctly with an IoU threshold, scored synthetic detections to obtain precision, recall, and VRU false-negative rate, and then exposed the dangerous night/occluded slice through per-condition recall analysis. The final recommendation was evidence-based and internally consistent, with SOTIF-style mitigations and a real-world comparison to the Uber ATG Tempe crash that tied the dossier back to an actual safety failure mode.

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