Decomposing the Neurons: Activation Sparsity via Mixture of Experts for Continual Test Time Adaptation
Please create and activate the following conda envrionment.
# It may take several minutes for conda to solve the environment
conda update conda
conda env create -f environment.yml
conda activate moase
- ViT as the backbone
Our source model is from timm, you can directly donwload it from the code.
Please load the source model from here
bash run_cifar10.sh # MoASE
Please load the source model from here
cd cifar
bash run_cifar100.sh # MoASE
For segmentation code, you can refer to cotta and SVDP. As for the source model, you can directly use Segformer trained on Cityscapes.
Please cite our work if you find it useful.
- CoTTA code is heavily used. official
- KATANA code is used for augmentation. official
- Robustbench official
- ImageNet-C Download