Why does 3D human-object reconstruction fail in the wild or get limited to a few object classes? A key missing piece is accurate 3D contact. InteractVLM (#CVPR2025) uses foundational models to infer contact on humans & objects, improving reconstruction from a single image. (1/10)

How can we infer 3D contact with limited 3D data? InteractVLM exploits foundational models—a VLM & localization model fine tuned to reason about contact. Given an image & prompt, the VLM outputs tokens for localization. But these models work in 2D, while contact is 3D. (4/10)

InteractVLM significantly outperforms prior work, both qualitatively and quantitatively, on in-the-wild 3D human (binary & semantic) contact prediction on the DAMON dataset. (6/10)

InteractVLM (#CVPR2025) is a great collaboration between Intelligent Systems, UvA Amsterdam and Inria. Authors: Sai Kumar Dwivedi, Dimitrije Antić, Shashank Tripathi, Omid Taheri, Cordelia Schmid, Michael Black and Dimitris Tzionas Code & models available at: interactvlm.is.tue.mpg.de (10/10)

If you’re at CVPR and interested in human-object interaction modeling, check out our poster!

🔥 New InteractVLM Models Released! (#CVPR2025) 🔹 Single Model for Joint Human-Object Contact 🔹 3D Human Contact trained on more data, now supports foot-ground contact 🔹 Direct Contact Estimation on Images (2D) 🔗 [Code] github.com/saidwivedi/Int…

neural oscillator of uncertain significance I just think about what bits need flipping and then point a laser pointer at the RAM to flip them.

🎯

"In CV we're slaves to the datasets. I think sometimes we're not creative or imaginative enough to make the datasets we ought to make, just those that are easy to make. I believe we have to go further and have annotations for 3D, even though it's really difficult." – Alan Yuille

"Our work is fully reproducible in the sense that it can be reproduced by simulating the laws of physics, starting from the initial configuration of the Universe."

Software engineers shouldn't fear being replaced by AI. They should fear being asked to maintain the sprawling mess of AI-generated legacy code their employer's systems will soon run on. Because that one will actually happen.

📢 Do u want to #Review for #CVPR2026, but u might not be in the reviewer pool? 📢 Do u also have prior publications at venues like #CVPR, #ECCV, #ICCV? 👉 Please DM me (or other ACs) with your: - First name - Last name - Email - OpenReview

Stochastic Gradient Descent

depth ambiguity