jung.vision

Computer Vision Research

[TEJASWINI @TEJASWINIMEDI13]

Excited to be at #WACV2025 🎉& share our work: "FAIR-TAT, an adversarial training framework to address adversarial fairness issues in DNN classifiers.
📍 Poster: March 3rd, from 11:15 AM
📄 https://t.co/T3ai8Ey8rD
Happy to connect and discuss 😃.

@margret_keuper @jung_vision

[Margret Keuper @margret_keuper]

If you are at #wacv2025 and interested in fair adversarial training, don't miss our poster presented by first author @TEJASWINIMEDI13 ! Joint work @dwsunima and @cvml_mpiinf

[Paul Gavrikov @PaulGavrikov]

Proud to announce that our paper "Can We Talk Models Into Seeing the World Differently?" was accepted at #ICLR2025 🇸🇬. This marks my last PhD paper, and we are honored that all 4 reviewers recommended acceptance, placing us in the top 6% of all submissions.

In our paper, we examine the formation of visual biases in large vision-language models (VLMs). Through the lens of the texture/shape bias, we show that LLMs significantly interact with the representation generated by dedicated vision encoders and utilize this finding to steer the VLM toward either extreme of a bias. All of our tested VLMs allowed to adjust the bias at inference time within some range by just utilizing natural language prompts! Or, in other words, we can simply talk a model into seeing the world differently. This is a key differentiator to traditional models, that will typically require finetuning or retraining to steer biases.

If this paper sounds familiar, then that is because you saw our abstract at the #CVPR2024 workshops :)

[Computer Vision and Machine Learning - MPI-INF @cvml_mpiinf]

"CosPGD: An efficient white-box adversarial attack for pixel-wise prediction tasks" by @shashankska, @jung_vision, @margret_keuper, in collaboration with @dwsunima.

https://t.co/TWGEqA3Bn3

[jung.vision @jung_vision]

Presenting our work on "Neural Architecture Design and Robustness: A Dataset" at #ICLR2023 with @jovita_lukasik. Check out our poster at 11:30 #132 and visit our project page: https://t.co/UGg2iXpO33

[Anna Kukleva @anna_kukleva_]

I’m excited to present our work with @MoritzBoehle at #ICLR. Check out our poster today at 16:30 “Temperature Schedules for self-supervised contrastive methods for long-tail data” https://t.co/Vu7FPy7Aey

[Julia Grabinski @JuliaGrabinski]

Living in the moment @eccvconf ☀️ with @jung_vision, Jovita Lukasik and @KeuperLabs

#ECCV22 #phdlife #TelAviv #conference #computervision

[ORTRA Conf & Events @ORTRALTD]

Poster presenters Jovita & Steffen are here from Germany, looking forward to their first ECCV conference! @eccvconf #eccv #eccv2022 #europeanconferenceoncomputervision #ai #computervision #computerengineering #telaviv #israel