Still have not seen a VLA do something like this. Enabled by optimisable 3D shape priors, 6 years ago!
edgarsucar.github.io/NodeSLAM/
@wkentaro_ @AjdDavison
Edgar Sucar
@SucarEdgar
- Come check out V-DPM @CVPR [Poster 25] 11:45 - 13:45 4D video reconstruction in the wild: code and models available 🤖 @EldarIsTyping @Oxford_VGG
- Fantastic results Stan! Great to see evidence of the usefulness of 3D representation/data for novel-view synthesis, when used in the right place.🍺 LagerNVS (CVPR 2026) 🍺 LagerNVS is a generalizable, feed-forward, real-time Novel View Synthesis network which - performs rendering in real time, - generalizes to in-the-wild data, - works with and without known source cameras, - sets a new state-of-the-art among
- Introducing V-DPM, for 4D reconstruction of in-the-wild videos. We build on top of VGGT, using Dynamic Point Maps for jointly representing 3D and motion. Joint work with: @EldarIsTyping , @LaiZihang , and Andrea Vedaldi. @Oxford_VGG. Check out the demo and code 👇
