This repository implements Lee et al., Hierarchical Novelty Detection for Visual Object Recognition, CVPR 2018 in PyTorch.
@inproceedings{lee2018hierarchical,
title={Hierarchical Novelty Detection for Visual Object Recognition},
author={Lee, Kibok and Lee, Kimin and Min, Kyle and Zhang, Yuting and Shin, Jinwoo and Lee, Honglak},
booktitle={CVPR},
year={2018}
}
You may download either raw images or ResNet-101 features. If you download ResNet-101 features, place them in datasets/{dataset}/
. ({dataset} = ImageNet, AWA2, CUB
)
datasets/ImageNet/train/
.datasets/ImageNet/train/n01440764/n01440764_18.JPEG
.datasets/ImageNet/known/
.datasets/ImageNet/novel/
.You do not have to download the files, but we provide the source of them for your reference.
Run sh scripts/preparation.sh {d}
. ({d} = imagenet_full, imagenet, awa2, cub
)
{d} = imagenet_full
if you have raw images{d} = imagenet
if you have ResNet-101 featuresOutput files are in taxonomy/{dataset}/
.
You can download pre-built taxonomies [here].
Run sh scripts/feature.sh {d}
. ({d} = imagenet, awa2, cub
)
Output files are in datasets/{dataset}/
.
If you have ResNet-101 features for ImageNet, skip this.
Run sh scripts/train.sh {d} {m}
. ({d} = imagenet, awa2, cub, {m} = relabel, td, loo, td+loo
)
Output files are in train/
.
You can download models reported in the paper [here].
datasets/{dataset}/balanced_order_{:d}.h5
datasets/{dataset}/relabels_{:d}.h5
train/
python build_taxonomy.py ImageNet detailed