Very happy to share that our (jung.vision,Margret Keuper,Keuper Labs) work, CosPGD: an efficient white-box adversarial attack for pixel-wise prediction tasks has been accepted at ICML 2024!
Pre-print arxiv: arxiv.org/pdf/2302.02213
Looking forward to ICML Conference in Vienna 😃

As an AC for #ECCV2024 , this is not NOT accurate……

This Google Scholar plugin is meaningfully accelerating my paper reviews. Clicking on a citation immediately pops up a helpful info box from Scholar, no more going back and forth to the references section, googling a title, reading the abstract, looping back, etc.

Computer Vision conference's acceptance criteria these days:
#CVPR2024 #eccc2024 #AI #ComputerVision

No 'Zero-Shot' Without Exponential Data: Pretraining Concept Frequency Determines Multimodal Model Performance

Web-crawled pretraining datasets underlie the impressive 'zero-shot' evaluation performance of multimodal models, such as CLIP for classification/retrieval and Stable-

Check out our new work on Generative Temporal Nursing for Longer Dynamic Video Synthesis 👇

yumengli007.github.io/VSTAR/

Thanks AK for sharing!

Thanks RSIP Vision for featuring our #ICLR2024 paper 'Adversarial Supervision Makes Layout-to-Image Diffusion Models Thrive'!
Check out our project page with code & models 🤗 yumengli007.github.io/ALDM

Joint work with Margret Keuper Dan Zhang Anna Khoreva at Bosch Center for Artificial Intelligence

Just about one year after the start of the research partnership between the MPI for Informatics and Google Deutschland, the VIA-Center Saarbücken is being expanded to add a research area “Vision and Language Models (VLMs)” under the direction of Prof. Dr. Bernt Schiele. ➡️📚sic.link/viavlm

Reflecting on an fantastic group retreat in the Austrian Alps last week. Engaging talks, insightful discussions, and a touch of recreation on the slopes! ⛷️🗻

🚀ALDM is accepted at #ICLR2024 ICLR 2024 !
We make layout-to-image diffusion models thrive, by integrating Adversarial Supervision to provide explicit alignment feedback & Multistep Unrolling for consistent adherence!
Code & Models are released🤗 yumengli007.github.io/ALDM

Proud to announce that our latest work was accepted at Transactions on Machine Learning Research!

If you’re interested in improving your CNNs robustness without specific training read more in our paper 👇

Do not let the #AIAct kill European AI research! An open letter of 150+ AI professors and PhDs asks to exempt research and open source from onerous requirements when sharing foundation models. docs.google.com/document/d/1u1…

Today, the MPI for Informatics presents its new #lightstage , dedicated to basic research in the field of visual computing. In terms of technical capabilities, it is among the leading facilities of its kind in the global research landscape🌎More: sic.link/lightstage VCAI - MPI for Informatics

Via Heiko Paulheim x.com/heikopaulheim/…

If you’re attending British Machine Vision Conference (BMVC), check out our paper “Divide&Bind” at oral session tomorrow at 4:30pm (UK) 👇

🚀 Thrilled to announce that I'll be presenting our Oral paper 'Divide & Bind' at #BMVC2023 tomorrow (21.11) at 16:30 (UK GMT time) 🗓️!

What to expect: Witness incredible text-to-image synthesis powered by improved Stable Diffusion, and the best part? No fine-tuning required! 🌟

Code for Divide & Bind has been released🤗Check it out!
Code: github.com/boschresearch/…
Project page: sites.google.com/view/divide-an…

Michael Brown #ICCV2023 reminds us that computer vision researchers should 'know their colour spaces'.
This tutorial in S06 should be called a 'masterclass'.
Glad to see a filled room - enough #ICCV2023 attendees are appreciating the fundamentals

Missed my poster #ICCV2023? Just ping me and let's discuss proper downsampling techniques for nice zebra representations! 🦓

Happy to be back Data and Web Science Group from today!