Excited to share our new paper (CVPR 2026 🚀): "MuViT: Multi-Resolution Vision Transformers for Learning Across Scales in Microscopy" which enables local predictions to use global context.
Great work led by @albertdm.bsky.social & another fun collab w @gioelelamanno.bsky.social! @scadsai.bsky.social
Since more than a decade, our group has been exploring chemically active droplets, which are exciting since their size can be controlled. Yet, we still do not fully understand this control even in the simplest models. Until a new preprint (arxiv.org/abs/2512.02542) spearheaded by postdoc Guido...
New #preprint: "Euler buckling on curved surfaces" arxiv.org/abs/2503.04303
@zhaoshh.bsky.social and I ask the question that Euler left open: How does an elastic line buckle within a (general) curved surface?
@mpipks.bsky.social @mpi-cbg.de @csbdresden.bsky.social
After nine years of development, meshoptimizer has reached its first major version, 1.0!
This release focuses on improvements in clusterization and simplification as well as stabilization. Here's a release announcement with more details on past, present and future; please RT!
meshoptimizer.org/v1
Turns out, there are fully opened LIDAR point scans of the whole Netherlands surface available geotiles.nl With up to 25 cm.
It was hard not to load our building and around it into #bigvolumebrowser (22 M points + Fire LUT for height)
Finally, the first public release of #BigVolumeBrowser, so after teasers, you can try it yourself. For details, please check the announcement post (1/2)
forum.image.sc/t/bigvolumeb...
Trackastra is part of the brand-new TrackMate v8.
It is the first deep-learning-based ("AI") drop-in replacement for TrackMate's regular LAP-Track algorithm.
Tracks out of the box for many types of datasets (bacteria, fluorescent nuclei, phase contrast cell culture etc), zero parameters to tune 🤖
I agree, is there a way to drop the name from the image mark (or maybe make multiple, one with the name big to the right of the logo?)
SynapseNet is a deep learning based software tool that automates the segmentation of vesicles, mitochondria, synaptic compartments, and the active zone. This is visualized in three panels. On the left, a section of an electron tomogram with a synaptic compartment densely filled with vesicles, which appear as round structures with dark boundary and light body in the image, is shwon. The top right panel shows the segmentation result of vesicles, visualized by masks with an individual color per vesicle and outlines for the segmented compartment (red) and active zone (blue). The bottom right panels shows a 3D rendering of the segmentation with vesicles shown as yellow spheres.
Are you studying synapses in electron microscopy? Tired of annotating vesicles? We have the tool for you! SynapseNet implements automatic segmentation and analysis of vesicles and other synaptic structures and has now been published:
www.molbiolcell.org/doi/full/10....
Are you rendering all points as emitters? Or do they have some form of occlusion/absorption/scattering?
Large images have to be broken into tiles both for training and inference with neural networks. The tile predictions then need to be merged to produce the final volume prediction.
Segment large images without tiling artifacts: sharing our work that should have been presented at ICCV in 2 weeks - the brilliant first author Elena can’t go because of visa issues.
The paper: arxiv.org/abs/2503.19545 1/🧵
Ah right, I have also seen some of the napari solution! I think this in general makes a lot of sense, and I'm still looking into how to make these kinds of volume interactions as easy in Blender as they are in the more scientific software, such as BigTrace and Napari 😅😋
This is separate from Blender, built on imglib2 in java (same framework as bigDataViewer, so also related to bigWarp), and for bigtrace all implemented by @ekatrukha.bsky.social
I think the smart thing it does for annotation is to extract the brightest point under the mouse as targeted point
This is the cool part of BigTrace, where this also has extensive semimanual and manual modes, while seeing your data in 3D. So this way, you can use the same engine to delete and fix traces.
Additionally i still want to make some automatic flags based on time info, but that will be more custom.
Automatic filament tracing in 3D data is now only a few clicks (or one macro) away! In the amazing #BigTrace FIJI plugin by @ekatrukha.bsky.social 🤩
My workflow will be to autotrace and then manually correct the autotraces, reducing the time for our tracing significantly!
Happy to share that ShapeEmbed has been accepted at @neuripsconf.bsky.social 🎉 SE is self-supervised framework to encode 2D contours from microscopy & natural images into a latent representation invariant to translation, scaling, rotation, reflection & point indexing
📄 arxiv.org/pdf/2507.01009 (1/N)
A cell finding its way through the matrix, imaged with @joycemeiri.bsky.social on LLS.
Amazing to see real-world medical applications of blender volume handling! 🤩🤩🤩
I gave a talk at Blender Conference yesterday, showing how biological volumes work, and how to visualize them in Blender 😋 youtu.be/WPajSWX730o?... #bcon25
I think it can be very useful to give a 'diorama' feel to a render and highlight certain complexity! But probably mostly for artistic purposes
How do you quantify cell structures which are deformed between different observations? Our preprint shows how to perform Simultaneous QUAntification of Structure and Structural Heterogeneity (SQUASSH) on 3D fluorescence microscopy data. (1/5)
www.biorxiv.org/content/10.1...
it was already fun to look around the documentation for cmap (cmap-docs.readthedocs.io)
but @grosoane.bsky.social just added a very cool feature to toggle between color vision deficiencies. Watch your perceptually uniform color maps gain some kinks 😂
as a mild deutan myself, I appreciate it! 🙏
Happy to officially introduce FilaBuster - a strategy for rapid, light-mediated intermediate filament disassembly. Compatible with multiple IF types, modular in design, and precise enough to induce localized filament disassembly in live cells.
www.biorxiv.org/content/10.1...
This is accompanied by a new set of tutorials showing exactly how you can start making beautiful microscopy visualizations in @blender.org!
Following along is also easy with the example OME-Zarr @openmicroscopy.org datasets 😄
Find the tutorials here: www.youtube.com/playlist?lis...
In the context of our @reviewcommons.org revision process, I'm happy to announce Microscopy Nodes v2.2.0!
This packs lots of new fun features, including new color management 🌈, clearer transparency handling 🫥, custom default settings 🔧 and more!
Preprint at doi.org/10.1101/2025...
New stable ilastik version 1.4.1 at www.ilastik.org/download!
Release Highlights:
- Support for OME-zarr (v0.4)
- New workflow - Trainable Domain Adaptation
- New object features based on Spherical Harmonics Feature, contributed by @grosoane.bsky.social.
- Native support for Apple Silicon
So excited to be a collaborator on this! We used @grosoane.bsky.social #microscopynodes to import tomo data to Blender before passing to Sensu for the beautiful animation.
It would only be appropriate to post about the #geometrynodes #microtubule here too. For a full introduction, I have prepared the world's least professional youtube video.
www.youtube.com/watch?v=AYUy...
My PhD work in the Köhler lab showing an involvement of a synaptonemal complex component in crossover regulation independently of synapsis is now published in NAR! 🎊 #meiosis #crossovers #C.elegans 🧬
Check it out here: academic.oup.com/nar/article/...
Together with Christian Tischer, 3D rendering of multi-res microscopy data was added to MoBIE FIJI plugin using #BigVolumeViewer.
Check any remote datasets (like @dudinlab.bsky.social maps of protists using ExM) in the highest quality.
For full details read forum.image.sc/t/mobie-plus...
