Improving Out-of-Distribution Detection via Epistemic Uncertainty Adversarial Training

Open in new window