How can we leverage successful pretraining techniques from transformers to improve purely convolutional networks? The answer is *Sparse Convolutions*! Let's see what happens when purely convolutional networks are pretrained with 1.28 million unlabeled images ... 1/7

SceneDreamer: Unbounded 3D Scene Generation from 2D Image Collections abs: arxiv.org/abs/2302.01330 project page: scene-dreamer.github.io

📢📢 📢 Introducing GaussCtrl!! ✅Easy control your Gaussian Splatting model with a text prompt! 🤩 ✅Remarkable multi-view consistency quality! w/ Jia-Wang Bian, Xinghui Li Guangrun Wang, Ian Reid, Philip Torr, and Victor Prisacariu 🌐gaussctrl.active.vision 📰arxiv.org/abs/2403.08733

Excited to announce the AI for Math Workshop at #ICML2024 ICML Conference! Join us for groundbreaking discussions on the intersection of AI and mathematics. 🤖🧮 📅 Workshop details: sites.google.com/view/ai4mathwo… 📜 Submit your pioneering work: sites.google.com/view/ai4mathwo… 🏆 Take on our

Our preprint on the astounding potential of Algebraic methods for IMO Geometry is out! 🚀 We developed the first system which outperforms an IMO gold medalist in Geometry! 🎉 Neuro-symbolic methods have a ton of potential. I'm super hyped! All the gory details below👇

New work with Lars Schaaf, Chen Lin, Guangrun Wang, and philip: We optimize neural networks to smoothly represent minimum energy paths and predict transition states for chemical reactions. Compared to the traditional approach, our method shows (i) improved resilience to

OMFA: One model to rule them all (virtual try-ons and offs)! Forget masks and multiple poses—just one pic, one pose, and voilà! Try-on, try-off, and super flexibility. It's practically magic for fashion tech. 🔮 - ChatGPT [Read more](arxiv.org/abs/2508.04559)

arxiv.org/abs/2510.01047 Authentic Discrete Diffusion Model