Shaping Visual Representations with Attributes for Few-Shot Learning

This code implements the Shaping Visual Representations with Attributes for Few-Shot Learning (ASL).

Citation

If you find our work useful, please consider citing our work using the bibtex:

@Article{chen2021asl,
	author  = {Chen, Haoxing and Li, Huaxiong and Li, Yaohui and Chen, Chunlin},
	title   = {Shaping Visual Representations with Attributes for Few-Shot Learning},
	journal = {arXiv preprint arXiv:2112.06398},
	year    = {2021},
}

Prerequisites

• Linux
• Python 3.7
• Pytorch 1.2
• Torchvision 0.4
• GPU + CUDA CuDNN

Datasets

You can download datasets automatically by adding --download when running the program. However, here we give steps to manually download datasets to prevent problems such as poor network connection: CUB:

1. Create the dir ASL/datasets/cub;
2. Download CUB_200_2011.tgz from here, and put the archive into ASL/datasets/cub;
3. Running the program with --download.

SUN:

1. Create the dir ASL/datasets/sun;
2. Download the archive of images from here, and put the archive into ASL/datasets/sun;
3. Download the archive of attributes from here, and put the archive into ASL/datasets/sun;
4. Running the program with --download.

Few-shot Classification

Download data and run on multiple GPUs with special settings:

python train.py --train-data [train_data] --test-data [test_data] --backbone [backbone] --num-shots [num_shots] --batch-tasks [batch_tasks] --train-tasks [train_tasks] --semantic-type [semantic_type] --multi-gpu --download

Run on CUB dataset, ResNet-12 backbone, 1-shot, single GPU

python train.py --train-data cub --test-data cub --backbone resnet12 --num-shots 1 --batch-tasks 4 --train-tasks 60000 --semantic-type class_attributes

Note that batch tasks are set to 4/1 when training 1-shot/5-shot tasks.

Our code is based on AGAM and TorchMeta.

Contacts

Please feel free to contact us if you have any problems.

Email: [email protected]