Landscape-Awareness for Geometric View Diffusion Model
For researchers working on 3D reconstruction and novel view synthesis, this work improves viewpoint estimation robustness under sparse views, though it is an incremental improvement over existing diffusion-based approaches.
The paper addresses the problem of accurate camera viewpoint estimation under sparse-view conditions, particularly in two-view scenarios. The proposed score-based method reshapes the optimization landscape to guide updates toward the ground-truth viewpoint, achieving competitive accuracy with higher sample-efficiency compared to existing methods.
Accurate camera viewpoint estimation under sparse-view conditions remains challenging, particularly in two-view scenarios. Recent approaches leverage diffusion models such as Zero123 to synthesize novel views conditioned on relative viewpoint, showing promising results when repurposed for viewpoint estimation via optimization with MSE loss. However, existing methods often suffer from nonconvex loss landscape with numerous local minima, making them sensitive to initialization and reliant on naive multistart strategies. We analyze these optimization challenges and visualize failure cases, showing that geometric ambiguities, such as symmetry and self-similarity, can mislead gradient-based updates toward incorrect viewpoints. To address these limitations, we propose a score-based method that reshapes the optimization landscape to guide updates toward the ground-truth viewpoint, followed by a refinement stage using a viewpoint-conditioned diffusion model. Experiments show that our method improves convergence, reduces reliance on brute-force sampling, and achieves competitive accuracy with higher sample-efficiency.