yes indeed! btw if one doesn't manage to finish the camera ready on time, will the abstract stay on openreview for all to see, or will it be removed?

🛡️ How do we secure AI systems as they grow more powerful? At the TDW, Mario Fritz mapped threats from LLMs & FMs and called for stronger safety benchmarks like CodeLMSec. ▶️ Watch the talk: youtu.be/RR1otLVODUk?...

Reminder for all the last minute registrations to ICLR and uploads for your WACV paper

At ICCV 2025, I am organizing two workshops: the LIMIT Workshop and the FOUND Workshop. ◆ LIMIT Workshop (19 Oct, PM): iccv2025-limit-workshop.limitlab.xyz ◆ FOUND Workshop (19 Oct, AM): iccv2025-found-workshop.limitlab.xyz We warmly invite you to attend at these workshops in ICCV 2025 Hawaii!

🗣️ Can generative AI speak more than English? Cees Snoek presents NeoBabel — a multilingual, open-source text-to-image model in 🇬🇧 🇨🇳 🇳🇱 🇫🇷 🇮🇳 🇮🇷, built for inclusivity and fairness. ▶️ Watch now: youtu.be/qUadAcykUiI?...

WACV'26 (R2,paper): 5 days + 2h. ICLR'26 (abs): 5 days + 2h. ICLR'26 (paper): 10 days + 2h. ECCV'26: 172 days.

I also discussed the DoS attack phenomenon with friends. We will have to find an answer for that by creating bars for submission (as sorry as I am for that). A place to look at might be the olympics with some preselection groups and strict doping rules. I don’t see how we can go on like this.

But scaling this is becoming a DoS attack on our reviewing institutions. It becomes a game of "can we fool the human referees?" 2/

I guess at the end of this semester, I can fairly confidently say that Artificial Idiocy (AI) is going to kill us all … 😂😭

📢 Job alert! #PlantSciJob We are opening a Professorship (W1-TT) in Plant Biochemistry at the ZMBP @unituebingen.bsky.social Apply by October 17th Please spread the word! More info 👇 uni-tuebingen.de/fakultaeten/...

Job alert! 🌸🏔️ The UW iSchool is hiring for *two* TT faculty positions, with a focus on Artificial Intelligence, broadly defined. It's pretty here! Applications due November 15th 2025 (priority deadline). Link: apply.interfolio.com/171020

Every lens leaves a blur signature—a hidden fingerprint in every photo. In our new #TPAMI paper, we show how to learn it fast (5 mins of capture!) with Lens Blur Fields ✨ With it, we can tell apart ‘identical’ phones by their optics, deblur images, and render realistic blurs.

WACV'26 (R2,reg): 4 days + 2h. WACV'26 (R2,paper): 11 days. ICLR'26 (abs): 11 days. ICLR'26 (paper): 16 days. ECCV'26: 178 days.

TIL that IEEE doesn't allow authors to use the 'Lena image'. For those who don't know, the Lena image is a crop of a sexually-explicit picture taken from Playboy that was used as the standard test image in computer vision papers for decades.

Strands #554 “One for the Trekkies” 🔵🔵🔵🔵 🟡🔵

Submission for the #WACV2026 Round 2 will open on OpenReview after reviews and decisions for Round 1 are released. We anticipate releasing Round 1 decisions by end of day anywhere on Earth on September 4th.

Strands #548 “Things are starting to take shape” 🔵🔵🔵🔵 🟡🔵🔵🔵

Strands #545 “No small matter” 💡🔵🔵🔵 🟡🔵🔵🔵

One unfortunate consequence of ... recent development ... is that we have more troubles getting PhD interns who are currently enrolled in PhD programs in the US: often they are afraid of not being able to come back to the US, and sometimes actually their universities are and don't let them go.

I just discovered this tree view. Why isn't this enabled by default? Only this really makes sense for complex threads.

Running experiments over the holidays 🤨

🚀 UTD is now fully released! Code ✅ Models ✅ 2M video descriptions ✅ Debiased splits for 12 datasets ✅ Everything you need to benchmark video models more fairly is now public: 🔗 github.com/ninatu/utd-p... 🎥 Let’s make video understanding actually about video understanding.

Update on hidden prompts in papers targeting LLM reviews: ICML 2025 PCs react. icml.cc/Conferences/...

Just a bit of cross-posting... 🚨 Nectar Track @ #GCPR2025 — Call for Submissions! 🧠📢 Have a top-tier paper in 24/25 at ICCV, ECCV, CVPR, NeurIPS, ICML, ICLR, PAMI, IJCV, etc.? Share your work with the vibrant GCPR community! 🗓️ Deadline: July 28, 2025 🔗 Instructions: dagm-gcpr.de/year/2025/su...

🚨 Big News! GCPR 2025 has received a record number of submissions this year!🎉 ⏰ Reminder: The supplementary deadline is tonight at 11:59 PM CEST. Good luck to all the authors wrapping things up! 💪✨ #GCPR2025 #ComputerVision Looking forward to the Fast Review Track submissions. Deadline: 1st July!

An amazing opportunity to share and discuss your top-tier work with an audience of top researchers!

New paper accepted to the Findings of #ACL2025: mRAKL: Multilingual Retrieval-Augmented Knowledge Graph Construction for Low-Resourced Languages. Work done in collaboration with folks from Apple (Min Li, @maartjeterhoeve.bsky.social , Saloni Potdar) and my advisor @schasins.bsky.social

Extended EPIC-SOUND paper was accepted at TPAMI arxiv.org/abs/2302.006... This follows ICASSP 2023 oral, extended for detection and further analysis epic-kitchens.github.io/epic-sounds/ work by @jaesunghuh.bsky.social Jacob Chalk @ekazakos.bsky.social @oxford-vgg.bsky.social @bristoluni.bsky.social

Replication Crisis xkcd.com/3117/

So… I have an idea! For those of us attending ICCV 2025 in Hawaii, I thought it would be very exciting and fun to collect an exotic dataset of tropical plants and animal species. I’m staying for vacations after the conference, and would like to combine it with the data collection. Who is joining?

Can MLLMs reason over images beyond the RGB domain? 🔍 We (@margretkeuper.bsky.social) explore In-Context Learning with VLMs for THz imaging to detect concealed threats, no training or finetuning in Smart Eyes for Silent Threats: VLMs and In-Context Learning for THz Imaging. arxiv.org/abs/2507.15576

Increasingly convinced the main damage cell phones have done to creativity isn't decreased attention spans or what not but the elimination of productive boredom. Inspiration comes from the mind filling the void inside. Hard to make art when you're constantly silencing the silence with "content."

🚨 @iccv.bsky.social Is inviting applications for our broadening participation call: iccv.thecvf.com/Conferences/... Deadline: Aug 20th 🌺

Eu calls for evaluator experts to counsel about how to evaluate models (can't belong to a training company, smart) digital digital-strategy.ec.europa.eu/en/news/comm... #safety #eticalAI #AI 🤖📈🧠 Tag people who should know or should pass the knowledge

🚨 Only 2.5 hours left! ⏳ Got a paper rejected from a peer-reviewed venue (e.g., BMCV, CVPR, ICCV, NeurIPS)? Submit your improved version + original reviews + rebuttal to the #GCPR2025 Fast Review Track! 📄📝 Give your work the second chance it deserves: 👉 www.dagm-gcpr.de/year/2025/su...

Do models really understand humans (ToM)? Can your model do it? Beat them at the neurips Competition. 🤖📈🧠

To everyone in the CV community, if you have never heard of AI slops, or you didn’t see the feature of LWT… In this case, please grab a coffee ☕️ and watch till the end. It is important for us and also has cat videos, if this helps.

I had a hard week and just realized I could really need one or more of those #icml puppies now 🐶🐶🐶

For any postdocs who are are looking to transition to independence and might be keen to join us here at @kingsioppn.bsky.social, King's Prize fellowship applications are open (deadline 27 Nov) 👇 See link for details, and feel free to reach out if you'd like to chat www.kcl.ac.uk/jobs/119505-...

I am giving a talk this morning at 10:40AM PST as part of the #ICML2025 Workshop on Assessing World Models. Title: "What Do Vision and Vision-Language Models Really Know About the World?" Come join us! www.worldmodelworkshop.org

Disinformation warriors are ‘grooming’ chatbots - on.ft.com/46hV6qW via @FT

EurIPS + ELLIS UnConference is coming to Copenhagen! 🇩🇰

Join us for 3rd Perception Test Workshop &Challenge @iccv.bsky.social #iccv2025 *NEW* this year: - 3 unified tracks - novel interpretability track - guest tracks: KiVA and Physics-IQ - 4 world-class speakers (see pic) Up to 50K in prizes sponsored by Google DeepMind 🧵 for details [1/4]

Video recordings from our workshop on Embodied Intelligence and tutorial on Robotics 101 @cvprconference.bsky.social are now up, just in time to catch up with things over the summer. Enjoy! #CVPR2025

I finally found the probably deepest reason why I don’t like LLMs…

I especially love how they’ve managed to make computers bad at math

Doing complicated math in greater volumes than ever before in order to arrive at incorrect answers to math problems a fourth grader knows how to do, it’s really something

EurIPS is coming! 📣 Mark your calendar for Dec. 2-7, 2025 in Copenhagen 📅 EurIPS is a community-organized conference where you can present accepted NeurIPS 2025 papers, endorsed by @neuripsconf.bsky.social and @nordicair.bsky.social and is co-developed by @ellis.eu eurips.cc

suspect it is possible for models to be much, much smaller than they are. Current models are so big because 1. They partially memorize the internet and 2. it seems to help with optimization. Any other arguments that they must be so big?