Rui Xiao

PhD Student @ TUM

[Rui Xiao @RuiXiao99]

Interested in VLM training with synthetic data? Wondering how to unlock fine-grained visual representations from captions?
I'm happy to present our work FLAIR @CVPR from 10:30 to 12:30 on 15th June, ExHall D Poster #368.
#CVPR2025

[Rui Xiao @RuiXiao99]

Huge credit to my wonderful friends & collaborators @SanghwanKim27, @xyongqin, @zeynepakata, Stephan Alaniz. And thanks to everyone I talked with at both conferences, it was great meeting you. See you at the next one!

[Rui Xiao @RuiXiao99]

Happy to have presented our work "FINER: MLLMs Hallucinate under Fine-grained Negative Queries" at #CVPR2026 as an oral (award candidate) and at #ICVSS2026
[Paper]: https://t.co/6z2H6W2mDw
[Code]: https://t.co/l5t6Vxq30H

[Rui Xiao @RuiXiao99]

We're fortunate (or maybe not) to live in an age where model capabilities advance this fast. Yet I still believe fine-grained alignment between vision and language is still not fully there — which would explain a lot of the problems we still see, and leaves room for future rs.

[Explainable Machine Learning @ExplainableML]

3/ (Oral)
FINER: MLLMs Hallucinate under Fine-grained Negative Queries
@RuiXiao99, @Sanghwa30296060, @xyongqin, @zeynepakata, Stephan Alaniz
[Paper]: https://t.co/6DNbIhGyyj
[Project]: https://t.co/8uxNSZRxAX
[Code]: https://t.co/JZYtP7GbJm

[Luca Eyring @LucaEyring]

Excited to be in San Diego at #NeurIPS2025 this week. We'll be presenting Noise Hypernetworks at Hall C,D,E #3605 Thu 4 Dec 11am!

If you're interested in Diffusion/Flow Matching or Reward Alignment, I'd love to chat! Feel free to DM me if you'd like to connect.

[Luca Eyring @LucaEyring]

Honored to receive one of the Google PhD Fellowships 2025 in ML Foundations for my research on "The Role of the Source Distribution in Generative Modeling: Beyond Fixed Gaussian Noise" :)
Grateful to @zeynepakata, Alexey, @TU_Muenchen, @Googleorg, and all my collaborators!

[Zeynep Akata @zeynepakata]

We are looking for PhD students. Please apply through the @ELLISforEurope and @MunichCenterML PhD programs!

[Rui Xiao @RuiXiao99]

RT @ExplainableML: We are proud of our PhD researcher Luca Eyring (@LucaEyring ) for his Google PhD Award 2025. Congratulations!🥳

[Explainable Machine Learning @ExplainableML]

🎓PhD Spotlight: Karsten Roth

Celebrate @confusezius, who defended his PhD on June 24th summa cum laude!

Karsten has been an @ELLISforEurope and IMPRS-IS PhD student since May 2021, supervised by both @zeynepakata and @OriolVinyalsML. His research has been centered around robust and effective deployment of (large) neural networks in the real world, with particular focus on a method and data-centric perspective for :

🦾 (Multimodal) model pretraining
🧠 Model generalization, reuse and transferability research
🎆 Continual (multimodal) training of such models.

During his PhD, Karsten interned @AIatMeta and @GoogleDeepMind, working on generalization in representation learning and large-scale multimodal pretraining techniques.

🏁 His next stop: @GoogleDeepMind in Zurich!

Join us in celebrating Karsten's achievements and wishing him the best for his future endeavors! 🥳

👇Checkout his selected publications in top-tier conferences such as NeurIPS, ICLR, CVPR or ICCV:

[Rui Xiao @RuiXiao99]

I'm happy to see the potential where we could unlock scalable & cunstomizable visual representations from generative data. Looking forward to further discussion at CVPR.

[Rui Xiao @RuiXiao99]

Interested in VLM training with synthetic data? Wondering how to unlock fine-grained visual representations from captions?
I'm happy to present our work FLAIR @CVPR from 10:30 to 12:30 on 15th June, ExHall D Poster #368.
#CVPR2025

[Rui Xiao @RuiXiao99]

6/
In terms of quantitative experiments, FLAIR show competative results in zero-shot standard, fine-grained, and long image-text retrieval, as well as decent segmentation performance. In zero-shot image classification, FLAIR lags behind CLIP-like models trained on billion-scale.