Building upon our previous work, Immiscible Diffusion (NeurIPS 2024), we present an improved version with deeper theoretical analysis and a significantly more efficient and effective implementation. This improved approach accelerates diffusion training across a wide range of

Very cool work from Yuxin Chen !

Our #CVPR2025 paper, "Bridging Viewpoint Gaps: Geometric Reasoning Boosts Semantic Correspondence," is being presented at Poster #90 in Exhall D today at 10:30 AM CDT! See firsthand how we've achieved ~40x faster performance with improved accuracy in semantic correspondence for rigid

Glad to share that our team effort CooPre received the Best Paper Award at the #CVPR25 DriveX Workshop (drivex-workshop.github.io)! Huge congratulations to Seth Z. Zhao for leading the way in making cooperative perception dramatically more data-efficient!

SVD is playing an increasingly important role in LLM. It has been widely used for model compression (e.g., our previous Dobi-SVD) and memory-efficient training (e.g., GaLore). Samir, Xiuyu, and Junxian take a creative step further by leveraging SVD to dynamically select a sparse

Check out our Sparse VideoGen! Xiuyu Li will be presenting—feel free to stop by and say hi!

Becoming an RL diehard in the past year and thinking about RL for most of my waking hours inadvertently taught me an important lesson about how to live my own life. One of the big concepts in RL is that you always want to be “on-policy”: instead of mimicking other people’s

Thank you Eric the Daydreamer and the Daydream team for using StreamDiffusion in building such an exciting platform for creators and builders! Together with あき先生(Aki)💧しずくラボ局長 and the rest of our team, we are continuously working to advance the open-source StreamDiffusion series for

July has been a big month for Viser! - Released v1.0.0😊 - We did some writing Some demos👇

Congratulations! Ryan Wang and Eigen AI

Very handy and well-engineered toolbox run with speed-of-light. Led by Jiahui Huang at NVIDIA AI. Project page: research.nvidia.com/labs/toronto-a…

[1/N] 🎥 We've made available a powerful spatial AI tool named ViPE: Video Pose Engine, to recover camera motion, intrinsics, and dense metric depth from casual videos! Running at 3–5 FPS, ViPE handles cinematic shots, dashcams, and even 360° panoramas. 🔗 research.nvidia.com/labs/toronto-a…