<ul data-eligibleForWebStory="false"><li>Generative models have difficulties in reliably evaluating sample quality for critical applications due to the concepts of fidelity and coverage.</li><li>To address this issue, two novel metrics, Clipped Density and Clipped Coverage, have been introduced to prevent out-of-distribution samples from biasing aggregated values.</li><li>These metrics exhibit linear score degradation as poor samples increase, making them easily interpretable as proportions of good samples.</li><li>Extensive experiments show that Clipped Density and Clipped Coverage outperform existing methods in terms of evaluating generative models in robustness, sensitivity, and interpretability.</li></ul>

Enhanced Generative Model Evaluation with Clipped Density and Coverage

Discover more