Déjà View: Looping Transformers for Multi-View 3D Reconstruction
Recent feed-forward 3D reconstruction transformers have scaled to over a billion parameters, following the broader trend of increasing model capacity in computer vision. Yet emerging evidence suggests that contiguous transformer layers often behave like repeated applications of similar operations, and multi-view reconstruction transformers refine their predictions progressively across decoder depth. We posit that model depth partially buys iteration, paid for inefficiently in unique parameters, and instead make that iteration explicit in architecture. Our model, DéjàView, applies a single looped transformer block recurrently to per-view features for K refinement steps. Trained once, it ex...
| Authors | Alessandro Burzio, Tobias Fischer, Sven Elflein et al. |
|---|---|
| Primary category | cs.CV |
| Published | 28 May 2026 |
| Last updated at source | 29 May 2026 |
Connected in the knowledge graph
- nvidia/dvltModel · cited by
Models and datasets referencing this paper
- nvidia/dvltModel
Silytics shows public metadata from arXiv and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.