Code for the paper "SelectiveNet: A Deep Neural Network with an Integrated Reject Option" ֿ To run an experiment run train.py --model_name [name] --baseline [name] --dataset [cifar_10/catsdogs/SVHN]
The calibration could be performed using the function post_calibration in selectivenet_utils.py