Multi-Granular Semantic Mining for Weakly Supervised Semantic Segmentation

This repository contains the PyTorch implementation for ACMMM 2022 Paper "Multi-Granular Semantic Mining for Weakly Supervised Semantic Segmentation" by Meijie Zhang, Jianwu Li, and Tianfei Zhou.

Prerequisites

We train our model using PyTorch 1.4.0 with four NVIDIA V100 GPUs with 32GB memory per card. Other Python modules can be installed by running

pip install -r requirements.txt

Training

Clone

git clone https://github.com/maeve07/HGNN

Dataset

Please download PASCAL VOC 2012 for training, and specify the path of the dataset.

Classification network

Please run python train.py for training the classification network.

Then generate the pseudo labels of the training set by running python gen_labels.py.

Segmentation network

We use Deeplab-v2 for the segmentation network with our generated pseudo labels. But most popular FCN-like segmentation networks can be used instead.

Citation

@article{zhou2021cascaded,
  title={Multi-Granular Semantic Mining for Weakly Supervised Semantic Segmentation},
  author={Zhang, Meijie and Li, Jianwu and Zhou, Tianfei},
  booktitle={ACM MM},
  year={2022}
}