Yugank completed Machine Learning for Automotive Safety

Built a perception validation dossier that correctly defined detection via IoU, scored a synthetic detections table with precision, recall, and VRU false-negative rate, and then stress-tested VRU recall across lighting, visibility, and rarity slices. The work shows strong safety-oriented evaluation habits: it distinguishes aggregate performance from vulnerable-road-user risk and surfaces the hardest conditions where the detector misses most often.

• Leakage-Free Detection Definition
• Honest Metric Evaluation
• Per-Condition Safety Triage
• Safety-Critical Risk Framing