🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

TOMORROW, during #CVPR2025 Session 2, get familiar with MONDAY (Mobile OS Navigation Task Dataset for Agents from YouTube), a large-scale dataset of 313K annotated frames, capturing diverse real-world mobile OS navigation across multiple platforms: cvpr.thecvf.com/virtual/2025/p…

🌍Presenting at #CVPR2025: MM-GRAPH–a new benchmark for multimodal graph learning! It extends beyond existing text-attributed graph benchmarks & offers a more comprehensive evaluation framework for multimodal graph neural networks (GNNs). Check it out! 🔗cvpr.thecvf.com/virtual/2025/p…

🥾🤖Let Humanoids Hike at #CVPR2025! Don't miss LEGO-H, a universal learning framework training humanoid robots to become expert hikers on complex trails by developing & integrating skills across all levels, embracing physical embodiment. 🔽 cvpr.thecvf.com/virtual/2025/p…

Our faculty and students are making 🌊waves🌊at #CVPR2025 — from lifelike humanoid robots to cutting-edge 3D reconstruction. Check out this overview of the work they’re presenting: cse.engin.umich.edu/stories/ten-pa…

🔥 Excited that 3D-MVP will be presented at #CVPR2025! It proposes a novel approach for 3D multi-view pretraining using masked autoencoders, leveraging Robotic View Transformer to improve generalization for downstream robotics tasks. 🔗cvpr.thecvf.com/virtual/2025/p…

📄 Catch this paper at #CVPR2025! It introduces a powerful multi-view diffusion model that jointly generates uncorrupted views from multi-view relationships — outperforming single-view and video methods on deblurring & super-res. 👉cvpr.thecvf.com/virtual/2025/p…

🧠📦 Introducing 3D-GRAND at #CVPR2025! A million-scale dataset to improve grounding and reduce hallucination in 3D large language models. Unlocking more reliable, spatially-aware AI. 🔗cvpr.thecvf.com/virtual/2025/p…

🎉 Excited to share that Prof. Stella Yu is playing a key role at #CVPR2025 — serving as Senior Area Chair, presenting cutting-edge research, and co-organizing workshops. Her leadership and contributions are helping shape the future of computer vision! 👏

Can we train mobile GUI agents just by watching YouTube videos? Yes; we present MONDAY at #CVPR2025 — a large-scale dataset with 313K automatically annotated frames from 20K real-world mobile OS tutorials (iOS + Android). ✅ No manual annotation ✅ Robust cross-platform

I'm excited to receive the NSF CAREER award! 🎉🎉 The funding will support our ongoing work to empower blind people to DIY personal assistive technology. Thanks to my students, mentors, collaborators, and U.S. National Science Foundation for making this possible. nsf.gov/awardsearch/sh…

👁️Check out Fast3R at #CVPR2025! 📍SUNDAY, June 15 @ 10:30 CDT -- Poster Session 5

Snapped a great moment at Michigan AI's #CVPR2025 meetup! 📸 Grateful for the chance to connect, share ideas, and catch up! Until next time. 👋

NEW ARTICLE: “AI generates data to help embodied agents ground language to 3D world.” The model trained on the synthetic dataset, 3D-GRAND, had higher grounding accuracy and hallucinated drastically less compared to their predecessors. bit.ly/3FCQcKn

We’re presenting 3D-MVP at CVPR poster #140 right now! #CVPR2025 #ComputerVision #MachineLearning #DeepLearning #3DVision #Robotics #AI #Research #PaperPresentation

Excited to share this important step towards understanding homophily and GNN performance in dynamic node classification! #AISTATS2025 Authors: Michael Ito, Danai Koutra, and Jenna Wiens: 📄arxiv.org/abs/2504.20421

Our newest newsletter just dropped!✨ Catch up on all the standout moments from this past spring — and see the amazing work our faculty and students have been doing: 📩 bit.ly/3SNjY1X

📢Newly released: Proud to share Michigan AI’s white paper, proposing a comprehensive framework for how we should evaluate real-world GenAI systems, emphasizing diverse, evolving inputs and holistic, dynamic, & ongoing assessment approaches. Now online: arxiv.org/abs/2504.16778