Home

Awesome

FRSKD

Official implementation for Refine Myself by Teaching Myself : Feature Refinement via Self-Knowledge Distillation (CVPR-2021)

Requirements

Classification Training

In this code, you can reproduce the experiment results of classification task in the paper. The datasets are all open-sourced, so it is easy to download. Example training settings are for ResNet18 on CIFAR-100. Detailed hyperparameter settings are enumerated in the paper.

python main.py --data_dir PATH_TO_DATASET \
--data CIFAR100 --batch_size 128 --alpha 2 --beta 100 \
--aux none --aux_lamb 0 --aug none --aug_a 0
python main.py --data_dir PATH_TO_DATASET \
--data CIFAR100 --batch_size 128 --alpha 2 --beta 100 \
--aux sla --aux_lamb 1 --aug none --aug_a 0
python main.py --data_dir PATH_TO_DATASET \
--data CIFAR100 --batch_size 128 --alpha 2 --beta 100 \
--aux none --aux_lamb 0 --aug mixup --aug_a 0.2
python main.py --data_dir PATH_TO_DATASET \
--data CIFAR100 --batch_size 128 --alpha 2 --beta 100 \
--aux none --aux_lamb 0 --aug cutmix --aug_a 1.0

Segmentation Training

You can reproduce the experiment results in FRSKD paper.

python main.py --data_dir PATH_TO_DATASET --batch_size 16 --alpha 1 --beta 50