[ECCV 2026] MetaView: Monocular Novel View Synthesis with Scale-Aware Implicit Geometry Priors

Yufei Cai¹  Xuesong Niu^2*  Hao Lu³  Kun Gai²  Kai Wu^2†  Guosheng Lin^1†

¹ Nanyang Technological University  ·  ² Kuaishou Technology  · 

³ The Hong Kong University of Science and Technology (Guangzhou)

^* Project lead    ^† Corresponding author

   

Overview

MetaView is a diffusion-based framework for high-fidelity monocular novel view synthesis that enables accurate rendering under large view changes from a single image.

Current generative novel view synthesis methods typically rely on restrictive explicit 3D reconstruction pipelines or use fully implicit scene modeling that suffers from scale drifting and poor geometry consistency. MetaView bridges this gap by combining implicit geometry modeling with minimal yet essential explicit 3D cues:

• Scale-Aware Implicit Geometry Priors: We extract hierarchical features and metric depth from a feed-forward geometry perception network (Depth Anything 3). These geometric signals are incorporated into the pretrained MM-DiT backbone (Qwen-Image-Edit) via non-invasive parallel attention layers, regularizing the spatial structure while preserving rich semantic knowledge.
• Metric Scale Anchoring via Modified RoPE: To overcome the scale drifting issue prevalent in fully implicit methods, we encode camera parameters into a modified Rotary Positional Encoding (PRoPE) and allocate an extra subspace for the z-axis. This explicitly injects metric scale cues, anchoring the generation to a consistent 3D metric space.

Given a single input image and a target camera pose, MetaView synthesizes the corresponding novel view with precise camera controllability, strong geometry consistency, and remarkable cross-domain generalization.