PANet: Few-Shot Image Semantic Segmentation with Prototype Alignment

This repo contains code for our ICCV 2019 paper PANet: Few-Shot Image Semantic Segmentation with Prototype Alignment.

Dependencies

• Python 3.6 +
• PyTorch 1.0.1
• torchvision 0.2.1
• NumPy, SciPy, PIL
• pycocotools
• sacred 0.7.5
• tqdm 4.32.2

Data Preparation for VOC Dataset

1. Download SegmentationClassAug, SegmentationObjectAug, ScribbleAugAuto from here and put them under VOCdevkit/VOC2012.

2. Download Segmentation from here and use it to replace VOCdevkit/VOC2012/ImageSets/Segmentation.

Usage

1. Download the ImageNet-pretrained weights of VGG16 network from torchvision: https://download.pytorch.org/models/vgg16-397923af.pth and put it under PANet/pretrained_model folder.

2. Change configuration via config.py, then train the model using python train.py or test the model using python test.py. You can use sacred features, e.g. python train.py with gpu_id=2.

Citation

Please consider citing our paper if the project helps your research. BibTeX reference is as follows.

@InProceedings{Wang_2019_ICCV,
author = {Wang, Kaixin and Liew, Jun Hao and Zou, Yingtian and Zhou, Daquan and Feng, Jiashi},
title = {PANet: Few-Shot Image Semantic Segmentation With Prototype Alignment},
booktitle = {The IEEE International Conference on Computer Vision (ICCV)},
month = {October},
year = {2019}
}