CVHub520 / X-AnyLabeling

Effortless data labeling with AI support from Segment Anything and other awesome models.
GNU General Public License v3.0
4.25k stars 485 forks source link

增加来自micro-sam项目,对光学/电子显微镜下细胞和细胞核做过fine-tune的模型 #186

Closed aiyou9 closed 10 months ago

aiyou9 commented 10 months ago

数据集来源

比赛:Weakly Supervised Cell Segmentation in Multi-modality High-Resolution Microscopy Images https://neurips22-cellseg.grand-challenge.org/ 第一名的Mean F1 做到了 0.9070

模型效果展示

https://computational-cell-analytics.github.io/micro-sam/micro_sam.html#which-model-should-i-choose

模型下载链接

https://github.com/computational-cell-analytics/micro-sam/blob/7651dda283473a05818a09d9d81646fa85468b5a/micro_sam/util.py#L45

CVHub520 commented 10 months ago

Hello, @aiyou9:

Thank you for your suggestion. We always appreciate it when users provide insights regarding relevant models that could be of assistance. Particularly, the inclusion of fine-tuned models for cell and nuclear imaging under optical/electronic microscopy from the micro-sam project seems promising.

If you already have refined versions of these models and would like to contribute to the improvement of the tool, I strongly recommend creating a Pull Request (PR) on GitHub. This would allow the contributors to the project to review and possibly incorporate your work into the existing framework.

Should you need any guidance regarding the process, feel free to ask for assistance. I'm here to help!

Best regards.

aiyou9 commented 10 months ago

ok I am try