Profile avatar
kristinmbranson.bsky.social
Researcher in machine learning and computer vision for science. Senior Group Leader at HHMI Janelia Research Campus. Supporter of DEIB in science and tech.
82 posts 1,003 followers 352 following
Prolific Poster
Conversation Starter

Every single philanthropic institution needs to do this rather than retracting right now. abcnews-go-com.cdn.ampproject.org/c/s/abcnews....

This is terrifying.

Exciting news! MegaSAM code is out🔥 & the updated Shape of Motion results with MegaSAM are really impressive! A year ago I didn't think we could make any progress on these videos: shape-of-motion.github.io/results.html Huge congrats to everyone involved and the community 🎉

I'm in a terrible Zoom meeting, so secretly I'm watching cute kea videos on reddit in the background: www.reddit.com/r/interestin... Here's a kea we ran into on the Kepler track in NZ last year trying to get into my backpack

Students and postdocs: interested in theoretical biophysics? Apply for @hhmijanelia.bsky.social’s second Junior Scientist Workshop on Theoretical Biophysics! Applications due in June for a work week of fantastic scientific interactions this October: www.janelia.org/you-janelia/...

We read "Large Concept Models" in our reading group. It models language at the sentence resolution - "concepts". We had lots of thoughts about it! You can read them in our new Janelia CVML blog post: janelia-cvml.github.io/blog/posts/L... #ML #AI #Janelia @adjavon.bsky.social @bogovicj.bsky.social

#Cellpose 3 paper now out. Not all images are perfect. Restore your images with Cellpose3 to get better segmentations, w/ @marius10p.bsky.social www.nature.com/articles/s41...

The anticipatory swiftness with which certain organizations go along with attacks on diversity and rewritings of history suggests that maybe some of the folks in power think that these are genuinely good changes.

Dr. Roger Airliner Young earned her PhD in 1940. Dr. Young studied #marine #biology and was the first African American woman to publish on #zoology in Science in 1924! marinesanctuary.org/blog/celebra... #28DaysOfBlackSTEMHistory #BlackHistoryMonth

Really excited to put together this #CVPR2025 workshop on "4D Vision: Modeling the Dynamic World" -- one of the most fascinating areas in computer vision today! We've invited incredible researchers who are leading fantastic work at various related fields. 4dvisionworkshop.github.io

If you are interested in fully documented fisheries, have a look at our recent work, AutoFish. A public instance segmentation and re-id dataset with cod-like fish on a conveyor. 450 fish 1,500 imgs 18,000 masks Accepted at MaCVi'25 @wacvconference.bsky.social Preprint: arxiv.org/abs/2501.03767

Announcing Diffusion Forcing Transformer (DFoT), our new video diffusion algorithm that generates ultra-long videos of 800+ frames. DFoT enables History Guidance, a simple add-on to any existing video diffusion models for a quality boost. Website: boyuan.space/history-guidance (1/7)

In honor of today’s International Day of Women and Girls in Science, The Transmitter spoke with eight women neuroscientists leading initiatives that strive to promote women in the field. www.thetransmitter.org/qa/how-eight... #WomenInScience #February11

Nice job, Transmitter, recognizing challenges for women in science and celebrating International Women in Science day, despite the current environment! Going to try out cleanBib!

Celebrating all those women who have, and continue to inspire me 🧡 #stem #womeninstem #science #InternationalDayofWomenandGirlsinScience

Do Large Language Models (LLMs) possess ecological knowledge? For example, can they do tasks such as: (1) predict the presence of species at a location (2) generate range maps (3) list critically endangered species (4) perform threat assessment (5) estimate species traits

Nice job, John's mom! Contacted my rep yesterday about the immigration uncertainty that is affecting my lab members, colleagues, and family. My representative is a Democrat, but I am hoping that having concrete examples of how this is affecting their constituents' lives can help them argue for us.

Very pleased to share our recent work, in which we use LLMs for automated discovery of interpretable models of animal behavior 🪰🐀🕵️‍♀️ that take the form of Python programs 🐍 See below for a summary of key results by @pcastr.bsky.social!

🚨Check it out! 🚨 Submissions are now open for the CellMap Segmentation Challenge! Signup here: cellmapchallenge.janelia.org

www.science.org/content/arti... Disappointing. The article also notes language changes l about the goals of the Hannah Gray, Freeman Hrabowski, and Gilliam programs.

If anyone has the resources to resist, its HHMI. Ending this program with no notice and literally scrubbing evidence it ever existed from the website is shocking and dosappointing

Over-Tokenized Transformer: Vocabulary is Generally Worth Scaling Hongzhi Huang, Defa Zhu, Banggu Wu, Yutao Zeng, Ya Wang, Qiyang Min, Xun Zhou tl;dr: increasing input vocabulary is always good, increasing output vocabularies is good for bigger models. arxiv.org/abs/2501.16975

There's a Korean soap opera on Netflix (When the Stars Gossip) which has *Drosophila courtship in space* as a major plot point (Ep 2). They have Michael Reiser & lab's flight arenas (you can see Janelia & MIchael's initials on the panels). Great way to learn about Drosophila courtship!

I have big news: @ellis.eu has launched its 2nd major research center, @ellisfinland.bsky.social! I have agreed to start as founding director & the first call for PI positions is open. This is a major opportunity for outstanding researchers, join us! ellisinstitute.fi/PI-recruit

For comparison: webcf.waybackmachine.org/web/20250103...

Exciting news! We’re thrilled to announce that tapir Yuna gave birth to a rare and endangered Malayan tapir calf Sunday night. The newborn, covered in distinctive white spots and stripes resembling a fuzzy walking watermelon, is only the second tapir born in our 120-year history.

Reinforcement learning agents should be able to improve upon behaviors seen during training. In practice, RL agents often struggle to generalize to new long-horizon behaviors. Our new paper studies *horizon generalization*, the degree to which RL algorithms generalize to reaching distant goals. 1/

I was supposed to give a talk in the NIH Neuroscience Seminar Series today, but it was cancelled because of Trump's executive orders. research.ninds.nih.gov/seminars-eve... Amongst the deluge, it seems that most Americans remain unaware that Trump is paralyzing the biomedical research community.

Nice clouds here in Virginia tonight

Health Programs Shutter Around the World After Trump Pauses Foreign Aid www.nytimes.com/2025/02/01/h...

I agree with the main message here 100%, but I'm not convinced that one method with unknown training protocol is proof of anything. I think a little bit of engineering on purely brute-force methods could give a factor of 50 increase in efficiency.

🚨 New paper 🚨 Benthic ecologist curious about using computer vision? 🤖 Computer scientist wondering how to help protect benthic biodiversity? 🦀 Our latest review bridges the gap between these fields! Check it out: www.sciencedirect.com/science/arti...

www.vox.com/politics/396... I've been disappointed in institutes for walking back efforts meant to improve DEI, but it now seems warranted, unfortunately

Sunrise in Joshua Tree

Kyle MacLachlan: How David Lynch Invented Me www.nytimes.com/2025/01/19/o... Thought this was pretty

Can you spot the difference between the cells on the left and right? Read on for the solution (and a preprint announcement)… 👇 1/6

MatchAnything: Universal Cross-Modality Image Matching with Large-Scale Pre-Training The authors propose a pre-training framework using synthetic cross-modal data to enhance LoFTR and RoMa for matching across medical imaging modalities like CT, MR, PET, and SPECT. zju3dv.github.io/MatchAnything/

I am shocked by the death of Felix Hill. He was one of the brightest minds of my generation. His last blog post on the stress of working in AI is very poignant. Apart from the emptiness of working mostly to make billionaires even richer, there's the intellectual emptiness of 'scale is all you need'