Maani Ghaffari
@GhaffariMaani
Roboticist; Assistant Professor of @NAMEatUM, University of Michigan Director of Computational Autonomy and Robotics Laboratory
Come and learn about state estimation, matrix Lie groups and exponential coordinates for robotics, and perception in Mobile Robotics as I teach in Winter 2022. All notes and files on Github: github.com/UMich-CURLY-te… Lectures on YouTube: youtube.com/watch?v=pH4Pkm…
🎉 Our paper "Registration beyond Points" was selected as a highlight at #ICCV2025! Kudos to the amazing team @joomeok98, Hyeonjae Gil, Junwoo Jang, and @GhaffariMaani Read thread to learn more 👇 Paper: arxiv.org/pdf/2507.17998 Code: github.com/joomeok/Grassm…
Registration beyond Points: General Affine Subspace Alignment via Geodesic Distance on Grassmann Manifold Jaeho Shin, Hyeonjae Gil, Junwoo Jang, @GhaffariMaani, @_ayoungk tl;dr: optimal cost function with geodesic distance on Grassmann manifold for line-plane registration…
Post RSS beach trip in LA with @JustinTzuYuan California is colder than Michigan! but today is nice.
Hanging out with program chair of @RoboticsSciSys and learning about the famous Hawaiian shirt history from @lucacarlone1 Wonderful conference and crowd.
Presented our poster at CVPR 2025 / HuMoGen this week in Nashville! Amazing to share ideas with the community and see our work in motion🤸 Thanks to the @humogen11384 organizers for making it all happen. 🔗 github.com/leob03/mdmp #CVPR2025 #HuMoGen #AI #MotionPrediction…
🚀 Our paper **MDMP** has been accepted at CVPR’25 - HuMoGen 🚀 We propose a multi-modal diffusion model that fuses textual action descriptions and 3D skeletal data to generate long-term human motion predictions, with interpretable uncertainty — paving the way for safer and…
And in the end, a tutorial by @CynthiaChienErh!
Join us for the 2nd workshop on Equivariant Vision: From Theory to Practice @CVPR on June 11, at Room 101C! Enjoy the exciting talks by @RobinSFWalters @GhaffariMaani @TacoCohen @_gabrielecesa_ @tesssmidt @vincesitzmann @twmitchel More details at: equivision.github.io
Happening now! @GhaffariMaani is talking about Lie Algebraic Neural Networks
Join us for the 2nd workshop on Equivariant Vision: From Theory to Practice @CVPR on June 11, at Room 101C! Enjoy the exciting talks by @RobinSFWalters @GhaffariMaani @TacoCohen @_gabrielecesa_ @tesssmidt @vincesitzmann @twmitchel More details at: equivision.github.io
this is a sick form of criticism. if you’ve a cogent argument to make about the merits of a paper make it. otherwise STFU about whether it was first-authored by an intern or not for the record interns are first authors on some of the best papers coming out of all tech companies
Dr. Yinshuang Xu @xu_yinshua86846, your defense was a masterpiece. It is rare that students rethink published papers just for the sake of the final version of thesis and defense presentation. This is what you did by introducing convolution in homogeneous spaces following Jean…
🚀 Our paper **MDMP** has been accepted at CVPR’25 - HuMoGen 🚀 We propose a multi-modal diffusion model that fuses textual action descriptions and 3D skeletal data to generate long-term human motion predictions, with interpretable uncertainty — paving the way for safer and…
Peer review process at scale
«Remember this video next time you're arguing with someone on the internet»
🚨 Just dropped: Nimble Dimples in action! 🌀💥 Our latest research from the Sareen Group shows how smart morphable skins can supercharge underwater agility. 🎥 Watch the video: youtu.be/De2VQYkdztM 📰 Full story: bit.ly/3K8y8Vd #FluidDynamics #SmartMorphableSkin
🚀 Gearing up for ICRA 2025 in Atlanta! Our Lunar Lab @ Georgia Tech is bringing the heat: 🌟 MI-HGNN & MS-HGNN for legged robot dynamics 🤖 STATE-NAV for humanoid traversability and navigation 🎙️ Prof. Lu Gan @ganlumomo speaks at two workshops: Foundation Models and…
Generative AI Foundations: Algorithms and Architectures jasour.github.io/generative-ai-…
In today's publication culture, most authors are after being SOTA, showing tables with 𝐛𝐨𝐥𝐝 numbers, and writing the minimum viable paper! The goal of a scientific paper should be to push the field forward with new intuition/insights on how to think about solving a problem.
I graduated 6 PhD students since last year! Ray Zhang -> Research Scientist at @ToyotaResearch @JustinTzuYuan -> Postdoc at MIT w/ @SangbaeMIT @CynthiaChienErh -> Applied Scientist at Amazon Robotics Tribhi Kathuria -> Lecturer at @UMRobotics @SangliTeng -> Postdoc at UC…
Vision and contact dynamics are both heavily influenced by geometry, so why do we treat them as separate problems? By combining vision with physics, "Vysics," each informs the other and we can generate accurate shape reconstructions despite major visual occlusions.