p2sam

KO | EN

gitlite — search

#javascript #python #ai #llm #machine-learning #docker #nodejs #cli #awesome-list #vue #pytorch #security #ios #mcp

(★ 11)

Part-aware Prompted Segment Anything Model for Adaptive Segmentation [TMLR 2025]

p2sam Latest Version Download

Download Latest Version (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# P2SAM
> Part-aware Prompted Segment Anything Model for Adaptive Segmentation\
> Chenhui Zhao and Liyue Shen\
> University of Michigan\
> [![TMLR](https://img.shields.io/badge/TMLR-accepted-success.svg)](https://openreview.net/forum?id=cCQKwd5MFP)&nbsp; [![arXiv](https://img.shields.io/badge/arXiv%20paper-2403.05433-b31b1b.svg)](https://arxiv.org/abs/2403.05433v2)&nbsp;

## Overview
<img src="https://github.com/Zch0414/p2sam/blob/master/figures/github.jpg" style="width:96%; max-width:800px; height:auto;" class="center">

We propose P2SAM, a training-free method that adapts a promptable segmentation model to new patients using only one-shot patient-specific data. 
P2SAM incorporates a **part-aware prompt mechanism** and a **distribution-based retrieval approach** to enhance generalization across:

- **Tasks:** P2SAM enhances generalization across different patient-specific segmentation tasks.
- **Models:** P2SAM can be plugged into various promptable segmentation models, such as SAM, fine-tuned SAM, and SAM 2.
- **Domains:** P2SAM performs effectively in both medical and natural image domains.

## Updates
- **(2025-07)** P2SAM's code has been released officially at [BioMedAI@UofM Repo](https://github.com/BioMed-AI-Lab-U-Michgan/p2sam).
- **(2025-05)** P2SAM is accepted by TMLR!
- **(2025-01)** Release P2SAM's SAM fine-tuning code and fine-tuned models.
- **(2025-01)** Release P2SAM code for adaptive nsclc segmentation.
- **(2025-01)** Release P2SAM code for endoscopy video segmentation.
- **(2024-07)** Release P2SAM code for personalized segmentation.

## Getting Started

Create an Python environment with:
```bash
conda env create -f environment.yaml
```

Prepare datasets with:
- [Datasets](data/README.md)

Prepare pre-trained and fine-tuned models with:
- [Models](pretrained_weights/README.md)

Fine-tune SAM on custom datasets (using SAM's iteratively training strategy) with:
- [Fine-tuning](fine_tuning/README.md)

Test P2SAM on out-of-distribution datasets with:
- [Testing](testing/README.md)

## Acknowledgement
This repository is built using the 
[DEiT](https://github.com/facebookresearch/deit/tree/main), 
[SAM](https://github.com/facebookresearch/segment-anything), 
[PerSAM](https://github.com/ZrrSkywalker/Personalize-SAM) 
repositories.

## Citation
If you find this repository helpful, please consider citing:
```bib
@article{zhaopart,
  title={Part-aware Prompted Segment Anything Model for Adaptive Segmentation},
  author={Zhao, Chenhui and Shen, Liyue},
  journal={Transactions on Machine Learning Research}
}
```

p2sam

p2sam Latest Version Download

Today's picks

Was this content helpful?

Recent Feedback

p2sam

p2sam Latest Version Download

Today's picks

Was this content helpful?

Recent Feedback

Download README