Stefan Hahmann (@stefanhahmann) 's Twitter Profile
Stefan Hahmann

@stefanhahmann

geoinformatics, bio image analysis, science & beyond

ID: 255464846

calendar_today21-02-2011 12:24:12

537 Tweet

268 Followers

436 Following

Stefan Hahmann (@stefanhahmann) 's Twitter Profile Photo

Mastodon Deep Lineage 0.0.1 released and uploaded to Fiji update site. Release Notes: github.com/mastodon-sc/ma… Update Site: sites.imagej.net/Mastodon-DeepL… Demonstration of functions: youtube.com/watch?v=774VE5…

Nature Methods (@naturemethods) 's Twitter Profile Photo

Out last week from the Tinevez and Olivo-Marin labs! JDLL: a library to run deep learning models on Java bioimage informatics platforms. This is sure to streamline deployment of DL models in bioimaging. nature.com/articles/s4159…

Out last week from the Tinevez and Olivo-Marin labs! JDLL: a library to run deep learning models on Java bioimage informatics platforms. This is sure to streamline deployment of DL models in bioimaging. nature.com/articles/s4159…
Stefan Hahmann (@stefanhahmann) 's Twitter Profile Photo

Mastodon Deep Lineage 0.1.1 released and uploaded to Fiji update site. Release Notes: github.com/mastodon-sc/ma… Update Site: sites.imagej.net/Mastodon-DeepL… Demonstration of functions: youtube.com/watch?v=kQakhh… Mastodon-DeepLineage now belongs to the managed update sites.

CEITEC (@ceitec_brno) 's Twitter Profile Photo

#HackPrague24 is underway 💻 this time in Prague! On 15-24 April we are focusing on improving Java-based biomedical image processing SW, in particular Fiji, BigDataViewer, and others 🔍 IT experts, common challenges, and innovation for science!

#HackPrague24 is underway 💻 this time in Prague! On 15-24 April we are focusing on improving Java-based biomedical image processing SW, in particular Fiji, BigDataViewer, and others 🔍
IT experts, common challenges, and innovation for science!
Robert Haase (@haesleinhuepf) 's Twitter Profile Photo

Stefan Hahmann Constantin Pape Fiji ChatGPT Carlos García If you need a second AI-opinion (😅): I asked gpt-4o and gpt-4-turbo a bit more systematically (n=25). Both seem to agree there are about 54 blobs. Such a quick test costs about $0.40 btw. 💸 haesleinhuepf.github.io/BioImageAnalys…

<a href="/stefanhahmann/">Stefan Hahmann</a> <a href="/cppape/">Constantin Pape</a> <a href="/FijiSc/">Fiji</a> <a href="/ChatGPTapp/">ChatGPT</a> <a href="/carlosg91018370/">Carlos García</a> If you need a second AI-opinion (😅): I asked gpt-4o and gpt-4-turbo a bit more systematically (n=25). Both seem to agree there are about 54 blobs. Such a quick test costs about $0.40 btw. 💸 
haesleinhuepf.github.io/BioImageAnalys…
Stefan Hahmann (@stefanhahmann) 's Twitter Profile Photo

Mastodon Deep Lineage 0.2.0 released and uploaded to Fiji update site. Release Notes: github.com/mastodon-sc/ma… Update Site: sites.imagej.net/Mastodon-DeepL… Demonstration of functions: youtube.com/watch?v=5GriMg…

Stefan Hahmann (@stefanhahmann) 's Twitter Profile Photo

I recently was involved in some installations of ELEPHANT server on Windows machines. I have written my learnings into this blog post: biapol.github.io/blog/stefan_ha…

I recently was involved in some installations of <a href="/ELEPHANT_track/">ELEPHANT</a> server on Windows machines.
I have written my learnings into this blog post: biapol.github.io/blog/stefan_ha…
Marcelo Zoccoler (@zoccolermarcelo) 's Twitter Profile Photo

🔀Tracking features can be clustered if represented via a Labels layer. We have a few notebooks in the repository teaching how to turn tracking features from Mastodon into a format compatible to napari-clusters-plotter here: github.com/BiAPoL/napari-…

Stefan Hahmann (@stefanhahmann) 's Twitter Profile Photo

Trying out the first (experimental!) release of Mastodon git (0.0.1), which can be used to collaboratively work on Mastodon tracking projects. Installation instructions and tutorials: github.com/mastodon-sc/ma… Psst, the video contains secret hints to the author of the tool.🧐

Johannes Soltwedel (@jm_mightypirate) 's Twitter Profile Photo

Finally got around to implement this: 💾I/O utilities for points and surfaces in 3D and 4D in napari🔬 - also partially supports saving feature so your data and tables stay together🤝 ➡️ github.com/jo-mueller/nap…

Robert Haase (@haesleinhuepf) 's Twitter Profile Photo

Kyle Douglass And one more shameless self-advertising read recommendation: If users learn basic programming, developers and users can meet in the middle minimizing total effort. (Fig lic CC BY) Johannes Soltwedel & yours truly. onlinelibrary.wiley.com/doi/10.1111/jm…

<a href="/kmdouglass/">Kyle Douglass</a> And one more shameless self-advertising read recommendation: If users learn basic programming, developers and users can meet in the middle minimizing total effort. (Fig lic CC BY) <a href="/jm_mightypirate/">Johannes Soltwedel</a> &amp; yours truly. onlinelibrary.wiley.com/doi/10.1111/jm…
Brendan Ashworth (@normconstant) 's Twitter Profile Photo

woke up this morning needing to digitize a map... so I spent 65 minutes drawing a USGS quad by hand, then 31 minutes with the AI vectorizer. here's an hour of #QGIS condensed into 60 seconds (top: AI, bottom: hand)

Johannes Soltwedel (@jm_mightypirate) 's Twitter Profile Photo

New blog post🔬💻! biapol.github.io/blog/johannes_… In our group's latest tutorial, I explain how you can train an Ultralytics YOLO object detection/classification 🔍 model for #bioimageanalysis directly from a dataset on an OMERO server openmicroscopy. Short summary 👇

New blog post🔬💻! biapol.github.io/blog/johannes_… In our group's latest tutorial, I explain how you can train an <a href="/ultralytics/">Ultralytics</a> YOLO object detection/classification 🔍 model for #bioimageanalysis directly from a dataset on an OMERO server <a href="/openmicroscopy/">openmicroscopy</a>. Short summary 👇
Johannes Soltwedel (@jm_mightypirate) 's Twitter Profile Photo

#OMERO is the go-to place to store and organize large amounts of image data. Its ability to annotate images - even collaboratively - makes a tedious task quite a bit easier. And for #YOLO, we luckily only need bounding boxes🖼️, anyway.

Johannes Soltwedel (@jm_mightypirate) 's Twitter Profile Photo

Which means, that we just need to clone a dataset, read out the annotations and reformat the bounding boxes and reformat them slightly (#OMERO ➡️#YOLO format) - and that's it, you're ready to train.

Which means, that we just need to clone a dataset, read out the annotations and reformat the bounding boxes and reformat them slightly (#OMERO ➡️#YOLO format) - and that's it, you're ready to train.
Johannes Soltwedel (@jm_mightypirate) 's Twitter Profile Photo

The results can be moved directly back to #OMERO using Erick Ratamero's amazing #ezomero . Super-helpful in one-shot detection &classification tasks (e.g., #smartmicroscopy). Data is available CC-BY 4.0: zenodo.org/records/133291… for you to reproduce. Looking forward to feedback!🌻

The results can be moved directly back to #OMERO using <a href="/erickratamero/">Erick Ratamero</a>'s amazing #ezomero . Super-helpful in one-shot detection &amp;classification tasks (e.g., #smartmicroscopy). Data is available CC-BY 4.0: zenodo.org/records/133291… for you to reproduce.
Looking forward to feedback!🌻
Robert Haase (@haesleinhuepf) 's Twitter Profile Photo

Assume you have access to the best LLM/AI for #BioImageAnalysis code generation: What would you ask it to code for you? 🙃 Please post your prompt below! Your input will become part of a public dataset shared under CC-BY 4.0, you will be acknowledged. #crowdsourcing Please RT. 🙏

Assume you have access to the best LLM/AI for #BioImageAnalysis code generation: What would you ask it to code for you? 🙃
Please post your prompt below! Your input will become part of a public dataset shared under CC-BY 4.0, you will be acknowledged. #crowdsourcing
Please RT. 🙏