Not All Synthetic Data Are Equal: Expert-Committee Audit Screening for Imbalanced Crash-Injury-Severity Prediction in Automated Driving Systems
Findings supporting a shift from quantity-oriented augmentation to credibility-aware sample acceptance for ADS safety prediction and risk governance support a shift from quantity-oriented augmentation to credibility-aware sample acceptance.
Ze-Wei Li, Qiao-Qiao Ren, Han Yang et al.
· 0 citations