NerfDiff: Single-image View Synthesis with NeRF-guided Distillation from 3D-aware Diffusion
This addresses the challenge of single-image view synthesis for computer vision applications, representing a novel hybrid approach rather than a foundational breakthrough.
The paper tackles the problem of generating novel views from a single image under severe occlusion, where existing methods produce blurry results. Their NerfDiff approach distills knowledge from a 3D-aware diffusion model into NeRF, significantly outperforming prior methods on datasets like ShapeNet, ABO, and Clevr3D.
Novel view synthesis from a single image requires inferring occluded regions of objects and scenes whilst simultaneously maintaining semantic and physical consistency with the input. Existing approaches condition neural radiance fields (NeRF) on local image features, projecting points to the input image plane, and aggregating 2D features to perform volume rendering. However, under severe occlusion, this projection fails to resolve uncertainty, resulting in blurry renderings that lack details. In this work, we propose NerfDiff, which addresses this issue by distilling the knowledge of a 3D-aware conditional diffusion model (CDM) into NeRF through synthesizing and refining a set of virtual views at test time. We further propose a novel NeRF-guided distillation algorithm that simultaneously generates 3D consistent virtual views from the CDM samples, and finetunes the NeRF based on the improved virtual views. Our approach significantly outperforms existing NeRF-based and geometry-free approaches on challenging datasets, including ShapeNet, ABO, and Clevr3D.