https://github.com/deeplearning-wisc/multi-label-ood

https://github.com/deeplearning-wisc/multi-label-ood

Last synced: 7 days ago
JSON representation

• Host: GitHub
• URL: https://github.com/deeplearning-wisc/multi-label-ood
• Owner: deeplearning-wisc
• Created: 2021-09-16T04:23:41.000Z (about 3 years ago)
• Default Branch: master
• Last Pushed: 2021-09-22T05:54:34.000Z (about 3 years ago)
• Last Synced: 2024-08-02T15:34:26.178Z (3 months ago)
• Language: Python
• Size: 3.65 MB
• Stars: 42
• Watchers: 1
• Forks: 5
• Open Issues: 3
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Can multi-label classification networks know what they don’t know? 

This is a [PyTorch](http://pytorch.org) implementation of [Can multi-label classification networks know what they don't know?](......) by Haoran Wang, Weitang Liu, Alex Bocchieri, Sharon Li. 

Code is modified from

[multilabel-ood](https://github.com/xksteven/multilabel-ood),

[ODIN](https://github.com/facebookresearch/odin),

[Outlier Exposure](https://github.com/hendrycks/outlier-exposure), and

[deep Mahalanobis

detector](https://github.com/pokaxpoka/deep_Mahalanobis_detector)

![teaser](demo_figs/teaser.png)

## Datasets

### In-distribution dataset

PASCAL-VOC: please download the dataset from

[this mirror](https://pjreddie.com/projects/pascal-voc-dataset-mirror/). Parsed labels for PASCAL-VOC are under ./dataset/pascal folder. Create the symlink to the location of Pascal dataset.

```

ln -s path/to/PASCALdataset Pascal

```

COCO: please download the MS-COCO 2014 dataset from [here](http://cocodataset.org/#download). Install the pycocotools to preprocess the dataset

```

pip3 install git+https://github.com/waleedka/coco.git#egg=pycocotools&subdirectory=PythonAPI

```

Preprocess the COCO dataset.

```

python3 utils/coco-preprocessing.py  path/to/coco-dataset

```

NUS-WIDE: please download the dataset from [here](https://lms.comp.nus.edu.sg/wp-content/uploads/2019/research/nuswide/NUS-WIDE.html). Parsed labels for NUS-WIDE are under ./dataset/nus-wide folder.

### Out-of-distribution dataset

OOD dataset can be downloaded

[here](https://drive.google.com/drive/folders/1BGMRQz3eB_npaGD46HC6K_uzt105HPRy?usp=sharing)

## Pre-trained models

Pre-trained models can be downloaded from

[here](https://drive.google.com/drive/folders/1ZfWB6vSYTK004j0bmfj6W0Xs6kwDTFX0?usp=sharing).

## Training the models

### Below are the examples on COCO dataset. 

Train the densenet model for COCO dataset

```

python3 train.py --arch densenet --dataset coco --save_dir ./saved_models/

```

Evaluate the trained model

```

python3 validate.py --arch densenet --dataset coco --load_path ./saved_models/

```

## OOD dection

To reproduce the JointEnergy score for COCO dataset, please run: 

```

python3 eval.py --arch densenet --dataset coco --ood_data imagenet --ood energy

--method sum 

```

To reproduce the scores for logit/msp/prob/lof/isol, please run:

```

python3 eval.py --arch densenet --dataset coco --ood_data imagenet --ood

logit/msp/prob/lof/isol/ --method max

```

To finetune the parameters for Odin and Mahalanobis, please run:

```

python3 fine_tune.py --arch densenet --dataset coco --ood odin/M --method max

```

After getting the best_T and best_noise, please run the evaluation:

```

python3 eval.py --arch densenet --dataset coco --ood_data imagenet --ood odin/M

--method max --T best_T --noise --best_noise

```

## OOD Detection Result

OOD detection performance comparison using JointEnergy vs. competitive

baselines.

![result](demo_figs/result_screenshot.png)

## Citation

     @article{wang2021canmulti,

          title={Can multi-label classification networks know what they don't know?},

          author={Wang, Haoran and Liu, Weitang and Bocchieri, Alex and Li, Yixuan},

          journal={Advances in Neural Information Processing Systems},

          year={2021}

     }