1 paper
Achim Hekler, Lukas Kuhn, Florian Buettner
Reliable uncertainty calibration is essential for safely deploying deep neural networks in high-stakes applications. Deep neural networks are known to exhibit systematic overconfid…