PSDNet

KO | EN

gitlite — search

#javascript #hacktoberfest #typescript #llm #android #machine-learning #deep-learning #linux #windows #awesome #cli #awesome-list #vue #ios

(★ 9)

[ICASSP 2025] First-frame Supervised Video Polyp Segmentation via Propagative and Semantic Dual-teacher Network

PSDNet Latest Version Download

Download Latest Version (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

<div align="center">
<h1>PSDNet</h1>
<h3>First-frame Supervised Video Polyp Segmentation via Propagative and Semantic Dual-teacher Network</h3>
 
<a href="https://scholar.google.com/citations?user=rU2JxLIAAAAJ&hl=en">Qiang Hu</a>1, Mei Liu2, Qiang Li1,&#8224;, <a href="https://scholar.google.com/citations?user=LwQcmgYAAAAJ&hl=en">Zhiwei Wang</a>1, &#8224;

1 WNLO, HUST, 2 HUST Tongji Medical College
 
(&#8224;: corresponding author)
</div>

## Overview
In this paper, we, for the first time, reduce the annotation cost to just ***a single frame*** per polyp video, regardless of the video's length. To this end, we introduce a new task, First-Frame Supervised Video Polyp Segmentation (FSVPS),
and propose a novel Propagative and Semantic Dual-Teacher Network (PSDNet). Specifically, PSDNet adopts a teacher-student framework but employs two distinct types of teachers: the propagative teacher and the semantic teacher. The propagative teacher is a universal object tracker that propagates the first-frame annotation to subsequent frames as pseudo labels. However, tracking errors may accumulate over time, gradually degrading the pseudo labels and misguiding the student model. To address this, we introduce the semantic teacher, an exponential moving average of the student model, which produces more stable and time-invariant pseudo labels. PSDNet merges the pseudo labels from both teachers using a carefully-designed back-propagation strategy. This strategy assesses the quality of the pseudo labels
by tracking them backward to the first frame. High-quality pseudo labels are more likely to spatially align with the firstframe annotation after this backward tracking, ensuring more accurate teacher-to-student knowledge transfer and improved segmentation performance.

## Models
### SUN-SEG Video Polyp Segmentation (VPS)

| Model | Backbone | Seen-Easy (Dice) | Seen-Hard (Dice) | Unseen-Easy (Dice) | Unseen-Hard (Dice) | Weights |
| :---- | :------: | :------: |:--:| :---------------: | :-----: | :-: |
| PSDNet | [PVT](https://drive.google.com/file/d/1U77oKKK_qik2C0fd7hSKiYG43UA25GgD/view)  | 0.900 | 0.860 | 0.798 | 0.806 |  [ckpts](https://github.com/Huster-Hq/PSDNet/releases/download/untagged-9a1b35db60def0c15641/PSDNet.pth) |

## Performance on [SUN-SEG](https://github.com/GewelsJI/VPS)

### 1. Quantitative Comparisons

 <img src="figs/results0.jpg"/>

### 2. Qualitative Comparisons

 <img src="figs/results1.jpg"/>

## Quick start
### - Preliminaries
- Python 3.8+
- PyTorch 1.9+ 
- TorchVision corresponding to the PyTorch version
- NVIDIA GPU + [CUDA](https://developer.nvidia.com/cuda-downloads)

#### 1. Install dependencies for SALI.
```bash
cd PSDNet
# Install other dependent packages
pip install -r requirements.txt

# Install cuda extensions for FA
cd lib/ops_align
python setup.py build develop
cd ../..
```

#### 2. Prepare the datasets for SALI.
Please refer to [PNS+](https://github.com/GewelsJI/VPS/blob/main/docs/DATA_DESCRIPTION.md) to get access to the SUN-SEG dataset, and download it to path `./datasets`. The path structure should be as follows:
```none
  SALI
  ├── datasets
  │   ├── SUN-SEG
  │   │   ├── TestEasyDataset
  │   │   │   ├── Seen
  │   │   │   ├── Unseen
  │   │   ├── TestHardDataset
  │   │   │   ├── Seen
  │   │   │   ├── Unseen
  │   │   ├── TrainDataset
  ```

### - Testing
```bash
python test_video.py
```

## Acknowledgments
Thanks [XMem](https://github.com/hkchengrex/XMem) for the implementation of an efficient universal video object segmentaion, which is used as the propagative teacher model in this work.

## Citation
If you find our paper and code useful in your research, please consider giving a star :star: and citation :pencil: :

```BibTeX
@inproceedings{zhou2025targeted,
  title={Targeted False Positive Synthesis via Detector-Guided Adversarial Diffusion Attacker for Robust Polyp Detection},
  author={Zhou, Quan and Luo, Gan and Hu, Qiang and Zhang, Qingyong and Zhang, Jinhua and Tian, Yinjiao and Li, Qiang and Wang, Zhiwei},
  booktitle={International Conference on Medical Image Computing and Computer-Assisted Intervention},
  pages={593--602},
  year={2025},
  organization={Springer}
}
```

PSDNet

PSDNet Latest Version Download

Today's picks

Was this content helpful?

Recent Feedback

PSDNet

PSDNet Latest Version Download

Today's picks

Was this content helpful?

Recent Feedback

Download README