3D reconstruction model that recovers geometry and camera poses from photos or video using a loop transformer.
Updated EU status checked Last shown
Sources: creator videos at the linked timestamps. EU status is information, not legal advice; see how we check.
Solver is a 3D reconstruction model described by a creator (AI Search) in a roundup of new 3D models. It reconstructs 3D geometry and camera poses from photos or video, using a loop transformer that refines its estimate incrementally.
According to that description, it supports streaming video input, so the reconstruction can be updated as new frames arrive. The creator said code and models were promised to be released within a week.
No official page, vendor, paper or license could be verified, so details such as benchmarks, licensing, model size and access are unknown.
Watch the moment: the exact second top creators show it. Videos load only when you press play, via youtube-nocookie.com.
Reconstructs 3D geometry and camera poses from photos or video using a loop transformer, refining incrementally and supporting streaming video. Code and models promised within a week.
Closest neighbours by meaning, not just by category.
Model · NVIDIA
NVIDIA's 117M-parameter looped transformer for multi-view 3D reconstruction of camera poses and dense geometry.
Model
Feed-forward 3D scene reconstruction model that outputs triangle primitives directly usable as meshes for simulation.
Research
Post-training method that refines feed-forward 3D reconstruction models (Pi3, DA3, VGGT) on unlabeled video.
Model
Feed-forward 3D Gaussian Splatting model that reconstructs scenes from arbitrary input views, with or without camera poses.
Research
Research method that reconstructs complete 4D (3D plus time) scenes from a single monocular video using Gaussian Splatting.
Model
Feed-forward system that turns an RGB image or video into editable, textured per-object 3D meshes in under a minute.