Strongly Pretrained Class Incremental Learning

This is code base for the following paper:

Class-Incremental Learning with Strong Pre-trained Models

Tz-Ying Wu, Gurumurthy Swaminathan, Zhizhong Li, Avinash Ravichandran, Nuno Vasconcelos, Rahul Bhotika, Stefano Soatto

Please read our paper for details!

Installation

To create a conda environment to run the project, simply run
conda env create -f CIL.yml.

Data

Create a soft link of the imagenet folder (the root folder that includes train/val image folders) at prepro/data/imagenet.

Experiments

800-40

bash scripts/getresults80040.sh -l layer4 -n 10  # for resnet10
bash scripts/getresults80040.sh -l layer4 -n 18  # for resnet18
bash scripts/getresults80040.sh -l fc -n 10  # for resnet10, fc-only
bash scripts/getresults80040.sh -l fc -n 18  # for resnet18, fc-only

500-50

bash scripts/getresults50050.sh -l layer4 -n 10
bash scripts/getresults50050.sh -l layer4 -n 18
bash scripts/getresults50050.sh -l fc -n 10
bash scripts/getresults50050.sh -l fc -n 18

Security

See CONTRIBUTING for more information.

License

This project is licensed under the Apache-2.0 License.