https://github.com/ChunmengLiu1/MECPformer

Implementation of "MECPformer: Multi-estimations Complementary Patch with CNN-Transformers for Weakly Supervised Semantic Segmentation"
https://github.com/ChunmengLiu1/MECPformer

Last synced: about 2 months ago
JSON representation

Implementation of "MECPformer: Multi-estimations Complementary Patch with CNN-Transformers for Weakly Supervised Semantic Segmentation"

• Host: GitHub
• URL: https://github.com/ChunmengLiu1/MECPformer
• Owner: ChunmengLiu1
• Created: 2023-02-13T09:19:33.000Z (almost 2 years ago)
• Default Branch: main
• Last Pushed: 2023-11-20T09:18:24.000Z (about 1 year ago)
• Last Synced: 2024-08-02T20:45:35.735Z (5 months ago)
• Language: Python
• Size: 2.68 MB
• Stars: 6
• Watchers: 2
• Forks: 0
• Open Issues: 1
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

• awesome-Transformers-For-Segmentation - MECPformer: Multi-estimations Complementary Patch with CNN-Transformers for Weakly Supervised Semantic Segmentation - **arXiv 2023** -[github](https://arxiv.org/pdf/2303.10689.pdf) (Papers / 2023)

README

        
# MECPformer

Official Implementation of the paper: [MECPformer: Multi-estimations Complementary Patch with CNN-Transformers for Weakly Supervised Semantic Segmentation](https://arxiv.org/pdf/2303.10689.pdf).

Accepted to Neural Computing and Applications.



## Abstract

The initial seed based on the convolutional neural network (CNN) for weakly supervised semantic segmentation always highlights the most

discriminative regions but fails to identify the global target information. Methods based on transformers have been proposed successively

benefiting from the advantage of capturing long-range feature representations. However, we observe a flaw regardless of the gifts based on

the transformer. Given a class, the initial seeds generated based on the transformer may invade regions belonging to other classes. Inspired

by the mentioned issues, we devise a simple yet effective method with Multi-estimations Complementary Patch (MECP) strategy and Adaptive

Conflict Module (ACM), dubbed MECPformer. Given an image, we manipulate it with the MECP strategy at different epochs, and the network mines and deeply fuses the semantic information at different levels. In addition, ACM adaptively removes conflicting pixels and exploits the network self-training capability to mine potential target information. Without bells and whistles, our MECPformer has reached new state-of-the-art 72.0% mIoU on the PASCAL VOC 2012 and 42.4% on MS COCO 2014 dataset. 

## Prerequisite

#### 1. install dependencies 

Ubuntu 20.04, with Python 3.6 and the following python dependencies.

```pip install -r requirements.txt```

#### 2. Download dataset 

Download [the PASCAL VOC 2012 development kit](http://host.robots.ox.ac.uk/pascal/VOC/voc2012).

#### 3. Download pretrained weights

Download [Conformer-S pretrained weights](https://drive.google.com/file/d/1qjLDy8MYU_TV2hspyYNCXeWrWho360qa/view?usp=share_link).

#### 4. Download saliency map

Download [saliency map](https://drive.google.com/file/d/1n7hVi8U2ylBMjz_bECsl_wSAmlRqnVr8/view?usp=share_link).

## Usage

#### 1. Run the run.sh script for training MECPformer in the initial seeds stage

```bash run.sh```

#### 2. Train semantic segmentation network

To train DeepLab-v2, we refer to [deeplab-pytorch](https://github.com/kazuto1011/deeplab-pytorch).

## Testing

#### Download our trained weights

| Stage       | Backbone   | Google drive | mIoU (%) |

|--------------|------------|--------------|--------------|

| Initial seeds | Conformer-S | [Weights](https://drive.google.com/file/d/1pYWyNjPTLVrZ5nlvipmgjRa_bDP6qSRE/view?usp=share_link)  | 66.6 |

| Final predict | ResNet101 | [Weights](https://drive.google.com/file/d/1thFitdq1MYq_BWKmNIlKZp1BCQzZFJQi/view?usp=share_link)  | 72.0 |

## Acknowledgements

This code is borrowed from [TransCAM](https://github.com/liruiwen/TransCAM), [CPN](https://github.com/Ferenas/CPN), and [deeplab-pytorch](https://github.com/kazuto1011/deeplab-pytorch).

## Citing MECPformer

If you use MECPformer in your research, please use the follwing entry:

```

@article{liu2023mecpformer,

  title={MECPformer: Multi-estimations Complementary Patch with CNN-Transformers for Weakly Supervised Semantic Segmentation},

  author={Liu, Chunmeng and Li, Guangyao and Shen, Yao and Wang, Ruiqi},

  journal={Neural Comput & Applic},

  year={2023}

}

```

or

```

Liu, C., Li, G., Shen, Y. et al. MECPformer: multi-estimations complementary patch with CNN-transformers for weakly supervised semantic segmentation. Neural Comput & Applic 35, 23249–23264 (2023). https://doi.org/10.1007/s00521-023-08816-2

```