dynamic_loss
Science Score: 57.0%
This score indicates how likely this project is to be science-related based on various indicators:
-
✓CITATION.cff file
Found CITATION.cff file -
✓codemeta.json file
Found codemeta.json file -
✓.zenodo.json file
Found .zenodo.json file -
✓DOI references
Found 1 DOI reference(s) in README -
○Academic publication links
-
○Academic email domains
-
○Institutional organization owner
-
○JOSS paper metadata
-
○Scientific vocabulary similarity
Low similarity (11.2%) to scientific vocabulary
Repository
Basic Info
- Host: GitHub
- Owner: jiangwenj02
- License: apache-2.0
- Language: Jupyter Notebook
- Default Branch: main
- Size: 2.44 MB
Statistics
- Stars: 4
- Watchers: 1
- Forks: 1
- Open Issues: 0
- Releases: 0
Metadata Files
README.md
Dynamic Loss for Robust Learning
This repository contains the source code for the research paper 'Dynamic Loss for Robust Learning,' built upon the mmclassification framework.
Requirements
- Python 3
- yapf (version ====0.40.1)
- PyTorch (version == 1.7.1)
- yaml
- higher(version == 0.2.1)
- mmcv-full (version == 1.5.0)
Installation
The installation process is similar to mmclassification. Please follow the same steps.
Training
python tools/train_meta.py configs/metadynamic/metadynamic_resnet32_cifar10_cor0.2_imb0.1.py
Visualization
After training, you can visualize the rank weight and margin.
python python tools/visualize_tools/vis_rank_margin.py --config configs/metadynamic/metadynamic_resnet32_cifar10_cor0.2_imb0.1.py --checkpoint work_dirs/metadynamic_resnet32_cifar10_cor0.2_imb0.1/latest.pth
The image will save in directory 'workdir/metadynamicresnet32cifar10cor0.2_imb0.1/'.
The label correct weight of each rank in each class.
Per class margin.
Citation
If you find Dynamic Loss useful, please cite the following paper
@ARTICLE{10238823,
author={Jiang, Shenwang and Li, Jianan and Zhang, Jizhou and Wang, Ying and Xu, Tingfa},
journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},
title={Dynamic Loss for Robust Learning},
year={2023},
volume={45},
number={12},
pages={14420-14434},
doi={10.1109/TPAMI.2023.3311636}}
Acknowledgement
The code is based on mmclassification and BalancedMetaSoftmax.Thanks for their great contributions on the computer vision community.
Owner
- Login: jiangwenj02
- Kind: user
- Repositories: 2
- Profile: https://github.com/jiangwenj02
Citation (CITATION.cff)
cff-version: 1.2.0 message: "If you use this software, please cite it as below." title: "OpenMMLab's Image Classification Toolbox and Benchmark" authors: - name: "MMClassification Contributors" version: 0.15.0 date-released: 2020-07-09 repository-code: "https://github.com/open-mmlab/mmclassification" license: Apache-2.0
GitHub Events
Total
- Watch event: 2
Last Year
- Watch event: 2
Dependencies
- pytorch/pytorch ${PYTORCH}-cuda${CUDA}-cudnn${CUDNN}-devel build
- pytorch/pytorch ${PYTORCH}-cuda${CUDA}-cudnn${CUDNN}-devel build
- docutils ==0.16.0
- m2r *
- myst-parser *
- sphinx ==4.0.2
- sphinx-copybutton *
- sphinx_markdown_tables *
- mmcv-full >=1.3.16,<=1.5.0
- albumentations >=0.3.2
- requests *
- mmcv >=1.3.16
- torch *
- torchvision *
- matplotlib *
- numpy *
- packaging *
- codecov * test
- flake8 * test
- interrogate * test
- isort ==4.3.21 test
- mmdet * test
- pytest * test
- xdoctest >=0.10.0 test
- yapf * test