LagerNVS: Latent Geometry for Fully Neural Real-time Novel View Synthesis

Stanislaw Szymanowicz, Minghao Chen, Jianyuan Wang, Christian Rupprecht, Andrea Vedaldi
3/20/2026
cs.CV

No official code found. Best implementation: facebookresearch/lagernvs (399 stars, last commit 2026-04-21).

Abstract

Recent work has shown that neural networks can perform 3D tasks such as Novel View Synthesis (NVS) without explicit 3D reconstruction. Even so, we argue that strong 3D inductive biases are still helpful in the design of such networks. We show this point by introducing LagerNVS, an encoder-decoder neural network for NVS that builds on `3D-aware' latent features. The encoder is initialized from a 3D reconstruction network pre-trained using explicit 3D supervision. This is paired with a lightweight decoder, and trained end-to-end with photometric losses. LagerNVS achieves state-of-the-art deterministic feed-forward Novel View Synthesis (including 31.4 PSNR on Re10k), with and without known cameras, renders in real time, generalizes to in-the-wild data, and can be paired with a diffusion decoder for generative extrapolation.

View on arXivView PDF

Code Implementations(1)

facebookresearch/lagernvsCommunity69%

Official code for "LagerNVS Latent Geometry for Fully Neural Real-time Novel View Synthesis" (CVPR 2026)

39926Python, Jupyter Notebook6 months agoNOASSERTION

Similar Papers

Autocodificadores Variacionales (VAE) Fundamentos Teóricos y Aplicaciones

82% match2023

Unsupervised Representation Learning from Pre-trained Diffusion Probabilistic Models

82% match1 repo2252022

Deep Learning for Case-Based Reasoning through Prototypes: A Neural Network that Explains Its Predictions

82% match1 repo772017

STAG: Structural Test-time Alignment of Gradients for Online Adaptation

82% match2024

SODAWideNet -- Salient Object Detection with an Attention augmented Wide Encoder Decoder network without ImageNet pre-training

81% match1 repo12023

Training the Untrainable: Introducing Inductive Bias via Representational Alignment

81% match2024

Cite this paper

@article{szymanowicz2026lagernvs,
  title  = {LagerNVS: Latent Geometry for Fully Neural Real-time Novel View Synthesis},
  author = {Stanislaw Szymanowicz and Minghao Chen and Jianyuan Wang and Christian Rupprecht and Andrea Vedaldi},
  year   = {2026},
  eprint = {2603.20176},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2603.20176v2}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2603.20176.svg)](https://codeofpaper.com/paper/2603.20176)

Discussion