This is the official repository of the paper:
DRAG: Dynamic Region-Aware GCN for Privacy-Leaking Image Detection
Guang Yang, Juan Cao, Qiang Sheng, Peng Qi, Xirong Li, and Jintao Li
To be in the Proceedings of the 36th AAAI Conference on Artificial Intelligence (AAAI 2022)
Preprint / Paper / Code / Short Video
We provide the main codes as well as example notebooks to utilize the codes.
The experimental datasets were from our previous papers. Refer to the paper and code for details. We provide a sample dataloader for the Image Privacy dataset in the dataset
.
python==3.6.8
torch==1.4.0
torchvision==0.5.0
The example scripts are in the channel_grouping_preprocess/ImagePrivacy/
.
Refer the codes to:
01. classification_pretraining.ipynb
.
02. channel_grouping.ipynb
.
03. channel_grouping_layer_pretraining.ipynb
.
Refer to the codes in DRAG_ImagePrivacy.ipynb
.
@article{yang2022drag,
title={DRAG: Dynamic Region-Aware GCN for Privacy-Leaking Image Detection},
author={Yang, Guang and Cao, Juan and Sheng, Qiang and Qi, Peng and Li, Xirong and Li, Jintao},
journal={arXiv preprint arXiv:2203.09121},
year={2022}
}