https://github.com/elijahcole/single-positive-multi-label

Multi-Label Learning from Single Positive Labels - CVPR 2021
https://github.com/elijahcole/single-positive-multi-label

computer-vision cvpr cvpr2021 deep-learning missing-labels multi-label-classification multilabel-classification

Last synced: 3 days ago
JSON representation

Multi-Label Learning from Single Positive Labels - CVPR 2021

• Host: GitHub
• URL: https://github.com/elijahcole/single-positive-multi-label
• Owner: elijahcole
• License: mit
• Created: 2021-05-24T00:22:32.000Z (over 3 years ago)
• Default Branch: main
• Last Pushed: 2023-11-21T01:41:35.000Z (12 months ago)
• Last Synced: 2024-11-05T14:47:00.475Z (5 days ago)
• Topics: computer-vision, cvpr, cvpr2021, deep-learning, missing-labels, multi-label-classification, multilabel-classification
• Language: Python
• Homepage: https://arxiv.org/abs/2106.09708
• Size: 25.8 MB
• Stars: 90
• Watchers: 8
• Forks: 18
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Multi-Label Learning from Single Positive Labels

Code to reproduce the main results in the paper [Multi-Label Learning from Single Positive Labels](https://arxiv.org/abs/2106.09708) (CVPR 2021). 

## Getting Started

See the `README.md` file in the `data` directory for instructions on downloading and setting up the datasets.

## Training a Model

To train and evaluate a model, run:

```

python train.py

```

## Selecting the Training Procedure

To generate different entries of the main table, modify the following parameters:

1. `dataset`: Which dataset to use.

1. `loss`: Which loss to use.

1. `train_mode`: Whether to (a) train a linear classifier on top of pre-extracted features, (b) train end-to-end, or (c) do (a) followed by (b).

1. `val_set_variant`: Whether to use a clean val set or a validation set where a single positive is observed for each image.

## Hyperparameter Search

As written, `train.py` will run a hyperparameter search over a few different learning rates and batch sizes, save the results for all runs, and report the best run. If desired, modify the code at the bottom of `train.py` to search over different parameter settings. 

**The `linear_init` mode searches over hyperparameters for the fine-tuning phase only.** The hyperparameters for the linear training phase are fixed. In particular, `linear_init_lr` and `linear_init_bsize` are set to the best learning rate and batch size from a `linear_fixed_features` hyperparameter search. 

## Misc

* The `requirements.txt` files was generated using the wonderful tool [pipreqs](https://github.com/bndr/pipreqs).

* Please feel free to get in touch / open an issue if anything is unclear. 

* In this paper we used only those images from NUSWIDE which were still publicly available when we re-crawled the dataset in 2020 using Namhyuk Ahn's [downloader](https://github.com/nmhkahn/NUS-WIDE-downloader). Following the instructions in `data/README.md` should yield the exact subset used for our experiments. 

## Reference  

If you find our work useful in your research please consider citing our paper:  

```latex

@inproceedings{cole2021multi,

  title={Multi-Label Learning from Single Positive Labels},

  author={Cole, Elijah and 

          Mac Aodha, Oisin and 

          Lorieul, Titouan and 

          Perona, Pietro and 

          Morris, Dan and 

          Jojic, Nebojsa},

  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},

  year={2021}

}

```