Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.
Mohan Zhou 7bcd4d5e29 | 3 years ago | |
---|---|---|
.. | ||
dataset | 3 years ago | |
models | 3 years ago | |
utils | 3 years ago | |
README.md | 3 years ago | |
requirements.txt | 3 years ago | |
train_index.py | 3 years ago |
This page provides basic tutorials about the usage of Look-into-Object for image classfication.
This code is tested with PyTorch 1.3.0 and torchvision 0.4.1.
python -m pip install -r requirements.txt
You can follow the Datasets Prepare
Section in DCL.
Note: The label_num
in annotations starts from 1 rather than 0.
Run train_index.py
to train CUB/STCAR/AIR.
Train with last stage and 3 positive images on CUB (LIO/ResNet-50 7x7):
python train_index.py --data CUB --stage 3 --num_positive 3
Feel free to open an issue if you encounter troubles.
If you use this codebase in your research, please cite our paper:
@InProceedings{Zhou_2020_CVPR,
author = {Zhou, Mohan and Bai, Yalong and Zhang, Wei and Zhao, Tiejun and Mei, Tao},
title = {Look-Into-Object: Self-Supervised Structure Modeling for Object Recognition},
booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2020}
}
该项目开源了一个基于图像中物体结构信息深度理解的图像识别、检测、分割算法框架。通过同一类别图像之间的共现信息首先基于弱监督的方式获得图像中物体的整体轮廓,然后基于自监督学习的方式,利用极坐标回归的损失函数来迫使神经网络对于目标物体的内部几何结构进行建模,从而辅助图像分类、检测以及分割任务来提升性能。
Jupyter Notebook Python Cuda Markdown C++ other
Dear OpenI User
Thank you for your continuous support to the Openl Qizhi Community AI Collaboration Platform. In order to protect your usage rights and ensure network security, we updated the Openl Qizhi Community AI Collaboration Platform Usage Agreement in January 2024. The updated agreement specifies that users are prohibited from using intranet penetration tools. After you click "Agree and continue", you can continue to use our services. Thank you for your cooperation and understanding.
For more agreement content, please refer to the《Openl Qizhi Community AI Collaboration Platform Usage Agreement》