Our work on calibrating deep neural networks is accepted at CVPR 2022

For details of this paper, please refer to : The Devil is in the Margin: Margin-based Label Smoothing for Network Calibration.

Applied Scientist

Related