search

CRBS / cdeep3m2

Plug-and-Play deep learning for image segmentation of large-scale microscopy data
Other
6 stars 3 forks source link

dataset of "SBEM, TEM Membranes (50789)" #24

Closed xinsheng44 closed 11 months ago

xinsheng44 commented 11 months ago

Hi there,

I hope you're doing well. I've been using your project and I'm quite impressed with its features. However, I encountered an issue while trying to segment my mitochondria data with "SBEM, TEM Membranes (50789)".

I am currently trying to train a proprietary domain segmentation model, and I have found that "SBEM, TEM Membranes (50789)" works very well for my sample data segmentation.

Since my labeled dataset is little, I would like to request dataset of "SBEM, TEM Membranes (50789)" . Can you make your training dataset publicly available?

I understand that you all are busy, but if you could spare some time to look into this, it would be greatly appreciated. Thank you so much for your hard work on this project!

Best regards, xinsheng

haberlmatt commented 11 months ago

Hi, if you are trying to segment Mitochondria, you can use this model: https://doi.org/10.7295/W9CDEEP3M50689 SBEM Mitochondria (denoised)

xinsheng44 commented 11 months ago

Hi, if you are trying to segment Mitochondria, you can use this model: https://doi.org/10.7295/W9CDEEP3M50689 SBEM Mitochondria (denoised)

Thank you for your reply.

I test most of the models on the Cdeep3M website and found that “https://doi.org/10.7295/W9CDEEP3M50789” is a better fit for my data because my data is 2D.

Here are the results of my tests.

https://doi.org/10.7295/W9CDEEP3M50789 image

https://doi.org/10.7295/W9CDEEP3M50689 image

But the effect is not as expected, so I want to get “https://doi.org/10.7295/W9CDEEP3M50789” corresponding training dataset.

So, could you open source the relevant data set for the corresponding model “https://doi.org/10.7295/W9CDEEP3M50789”?

Thanks!