Ich habe kein Verständnis für die massive Unterstützung der Interessen von RWE durch Armin Laschet und die Landesregierung NRW. Der Hambacher Forst braucht jetzt eine politische Lösung! #StopptdenWahnsinn #kohlefrei wwf.de/stoppt-das

Many lessons for self-supervised representation learning: arxiv.org/abs/1901.09005. The best case halves the gap between previous SOTA and supervised training on ImageNet! Code: github.com/google/revisit…. Work with Alexander Kolesnikov and Xiaohua Zhai at Google Brain Zürich Google AI.

With the help of the ear loops I tore off a spare mask, I'm now the owner of what I think is the first silenced #cvpr2024 lanyard

I very much like the large art exhibit at #CVPR2024 ❤️ To bad it cut down my time for posters even more 😅

#CVPR2024 well played...

That's an interesting approach to problem solving #ChatGPT

Fine-Tuning Image-Conditional Diffusion Models is Easier than You Think discuss: huggingface.co/papers/2409.11… Recent work showed that large diffusion models can be reused as highly precise monocular depth estimators by casting depth estimation as an image-conditional image

Check out our work on fine-tuning of image-conditional diffusion models for depth and normal estimation. Widely used diffusion models can be improved with single-step inference and task-specific fine-tuning, allowing us to gain better accuracy while being 200x faster!⚡ 🧵(1/6)

Image segmentation doesn’t have to be rocket science. 🚀 Why build a rocket engine full of bolted-on subsystems when one elegant unit does the job? 💡 That’s what we did for segmentation. ✅ Meet the Encoder-only Mask Transformer (EoMT): tue-mps.github.io/eomt (CVPR 2025) (1/6)

I'll be presenting "DINO in the Room (DITR)", the winning method of the ScanNet++ 3D semantic segmentation challenge, tomorrow at CVPR at 10 a.m. in Room 211. Project page: visualcomputinginstitute.github.io/DITR/

Drinking coffee with soy milk at work... A whole new world just opened :o <3

Woohoo, spent 4 hours on a #presentation everyone told me not to put too much effort into x) Finished at 3 AM :x Lots of #animations though

Dear me. When did it get to be 1am?!

Thinking all your experiments were wrong... 2 days prior to the submission #deadline! Haven't had such a bad feeling in a long time!

So many typos :o I'm totally taking the negative "logatirhm" of my probability distributions :D

up and running for 24 hours now. Let's hope it was worth it!

Someone about me: "The guy's #nickname is #Unicorn, who could be more gay?"

How do you mean: "The wall is not a portal!" ? #confused #StupidWall #ERW2014 #RobotMarathon