BFC-BL: Few-Shot Classification and Segmentation combining Bi-directional Feature Correlation and Boundary constraint

Environmnet installation

Make sure to replace YOURCONDADIR in the installation path with your conda dir, e.g., ~/anaconda3

Set up the following basic environment:

• Ubuntu 18.04或Windows 10（带有Anaconda/Miniconda）
• Python 3.7
• CUDA 11.0

conda create -n bfc-bl python==3.7
conda activate bfc-bl

The required packages are as follows：

• pytorch==1.7.0
• torchvision==0.8.1
• cudatoolkit==11.0.3
• pytorch-lightning==1.3.8
• einops==0.3.0

Datasets

• PASCAL VOC 2012 and SBD

    YOUR_DATASET_DIR/
    ├── VOC2012/
    │   ├── Annotations/
    │   ├── JPEGImages/
    │   ├── ...

We follow the dataset protocol of ASNet, HSNet and PFENet.

Related works

Our project refers to and heavily borrows some the codes from the following repos:

• [PANet]: Wang et al., Few-shot image semantic segmentation with prototype alignment, ICCV'19.
• [PFENet]: Tian et al., Prior guided feature enrichment network for few-shot segmentation, TPAMI'20.
• [HSNet]: Min et al., Hypercorrelation squeeze for few-shot segmentation, ICCV'21.
• [ASNet]: Kang et al., Integrative Few-Shot Learning for Classification and Segmentation, CVPR'22 =======