Princeton Visual AI lab
@VisualAILab
Research on computer vision, AI/human interaction, vision+language, AI fairness+transparency. PI @orussakovsky. #PrincetonCS
You might like
Text-to-image (T2I) models excel in semantic alignment, yet numeracy—generating exact numbers of objects—remains challenging. We propose D2D: Detector-to-Differentiable Critic for Improved Numeracy in Text-to-Image Generation to overcome this. arxiv.org/abs/2510.19278 (1/7)
Prof. Johannes Lutzeyer (@JLutzeyer) and I are co-recruiting a fully funded PhD student at the Department of Computer Science, École Polytechnique in France, to start in September 2026. Details below or at l-yezhu.github.io/assets/PhD_Hir… #PhDhiring #AI
Text-to-image (T2I) models can generate rich supervision for visual learning but generating subtle distinctions still remains challenging. Fine-tuning helps, but too much tuning → overfitting and loss of diversity. How do we preserve fidelity without sacrificing diversity (1/8)
🚀 Our paper “The Impact of Coreset Selection on Spurious Correlations and Group Robustness” (with William Yang, @polkirichenko , Lydia Liu & @orussakovsky) is accepted to #NeurIPS2025 Datasets & Benchmarks! 🎉 @NeurIPSConf 🧵 Key findings 👇 [Paper: openreview.net/pdf?id=t7QJDaQ…]
![Amaya_Dharmasir's tweet image. 🚀 Our paper “The Impact of Coreset Selection on Spurious Correlations and Group Robustness” (with William Yang, @polkirichenko , Lydia Liu & @orussakovsky) is accepted to #NeurIPS2025 Datasets & Benchmarks! 🎉 @NeurIPSConf
🧵 Key findings 👇
[Paper: openreview.net/pdf?id=t7QJDaQ…]](https://pbs.twimg.com/media/G4iUMLrXEAArwPU.jpg)
Cornell (NYC and Ithaca) is recruiting AI postdocs, apply by Nov 20, 2025! If you're interested in working with me on technical approaches to responsible AI (e.g., personalization, fairness), please email me. academicjobsonline.org/ajo/jobs/30971
I'm at #ICCV2025 this week🌸 - We will be hosting Curated Data for Efficient Learning workshop (w @GCazenavette) 🗓️ Oct. 20, 10:30-10:35 AM 📍Room 304A 🔗 curateddata.github.io - I will be presenting our recent work on "Where Is Motion From? Scalable Motion Attribution for…
Our Perception Test kickoff is happening now in Ballroom B (top floor) from @joaocarreira! #ICCV2025
Salma Abdel Magid and Tiwalayo Eisape join the computer science department as Presidential Postdoctoral Research Fellows. The fellowship recognizes and supports 15 outstanding scholars from disciplines across the university. bit.ly/4oauk9E
Excited to share that I've joined @Polytechnique as a tenure-track assistant professor in Computer Science @LIX_lab. Huge thanks to my dear Princeton friends @VisualAILab and mentor @orussakovsky💫 Looking forward to new adventures with colleagues (@mvazirg @JLutzeyer) in France!
How to adapt VLMs into robot policies without losing their multimodal generalization capability? We propose VLM2VLA - we treat robot actions as language and apply LoRA tuning. This preserves core VLM skills while enabling strong zero-shot generalization to new tasks & languages.
Why don’t VLAs generalize as well as their VLM counterparts? One culprit: catastrophic forgetting during fine-tuning. 🧠 We introduce VLM2VLA: a training paradigm that preserves the VLM capabilities while teaching robotic control. vlm2vla.github.io 🧵
Why don’t VLAs generalize as well as their VLM counterparts? One culprit: catastrophic forgetting during fine-tuning. 🧠 We introduce VLM2VLA: a training paradigm that preserves the VLM capabilities while teaching robotic control. vlm2vla.github.io 🧵
Does how we talk about AI matter? Yes, yes it does! In our #Cogsci2025 paper, we explore how different messages affect what people believe about AI systems. See 🧵for more! Paper: allisonchen.us/assets/portray…
Have you ever felt that AI fairness was too strict, enforcing fairness when it didn’t seem necessary? How about too narrow, missing a wide range of important harms? We argue that the way to address both of these critiques is to discriminate more 🧵
Commencement 🐯🎓🎉
📢 I successfully defended my PhD dissertation today! Huge thanks to my committee @orussakovsky @andresmh @jennwvaughan @QVeraLiao @parastooabtahi & everyone who supported me ❤️ 📢 Next I'll be joining Apple as a research scientist in the Responsible AI team led by Jeff Bigham!
First day at NVIDIA as a research scientist intern:) If you’re in the Bay and want to chat about VLM or video gen related research, especially around compressibility, compositionality, or controllability, feel free to DM me!
📢 I successfully defended my PhD dissertation today! Huge thanks to my committee @orussakovsky @andresmh @jennwvaughan @QVeraLiao @parastooabtahi & everyone who supported me ❤️ 📢 Next I'll be joining Apple as a research scientist in the Responsible AI team led by Jeff Bigham!
Introducing COMPACT: COMPositional Atomic-to-complex Visual Capability Tuning, a data-efficient approach to improve multimodal models on complex visual tasks without scaling data volume. 📦 arxiv.org/abs/2504.21850 1/10
Check out Indu Panigrahi’s LBW at #CHI2025: “Interactivity x Explainability: Toward Understanding How Interactivity Can Improve Computer Vision Explanations.” 🔗 Project Page: ind1010.github.io/interactive_XAI 📄 Extended Abstract: arxiv.org/abs/2504.10745
"Fostering Appropriate Reliance on LLMs" received an Honorable Mention at #CHI2025! This work is also the last chapter of my dissertation, so the recognition feels more special 🏅🎓😊 🎉 to the team @jennwvaughan @QVeraLiao @TaniaLombrozo @orussakovsky
Appropriate reliance is key to safe and successful user interactions with LLMs. But what shapes user reliance on LLMs, and how can we foster appropriate reliance? In our #CHI2025 paper, we explore these questions through two user studies. 1/7
United States Trends
- 1. #NXXT_Earnings N/A
- 2. Good Friday 46.7K posts
- 3. #FanCashDropPromotion N/A
- 4. Summer Walker 21.2K posts
- 5. #FridayVibes 3,868 posts
- 6. #GringosVenezuelaNoSeToca N/A
- 7. Wale 35.6K posts
- 8. #FinallyOverIt 6,899 posts
- 9. Happy Friyay 1,160 posts
- 10. RED Friday 2,513 posts
- 11. Go Girl 25.7K posts
- 12. Meek 6,189 posts
- 13. Saylor 29.8K posts
- 14. $BTC 119K posts
- 15. Robbed You 4,625 posts
- 16. Bubba 8,739 posts
- 17. SONIC RACING 2,304 posts
- 18. Monaleo 2,300 posts
- 19. For Christ 22.4K posts
- 20. Bihar 351K posts
You might like
-
Silvio Savarese
@silviocinguetta -
WACV
@wacv_official -
Angjoo Kanazawa
@akanazawa -
Jun-Yan Zhu
@junyanz89 -
Raquel Urtasun
@RaquelUrtasun -
Georgia Gkioxari
@georgiagkioxari -
Judy Hoffman
@judyfhoffman -
Princeton Computer Science
@PrincetonCS -
Kate Saenko
@kate_saenko_ -
Phillip Isola
@phillip_isola -
Vittorio Ferrari
@VittoFerrariCV -
Jiajun Wu
@jiajunwu_cs -
Noah Snavely
@Jimantha -
Devi Parikh
@deviparikh -
Shuran Song
@SongShuran
Something went wrong.
Something went wrong.