English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Novel view synthesis from a single image requires inferring occluded regionsof objects and scenes whilst simultaneously maintaining semantic and physicalconsistency with the input. Existing approaches condition neural radiancefields (NeRF) on local image features, projecting points to the input imageplane, and aggregating 2D features to perform volume rendering. However, undersevere occlusion, this projection fails to resolve uncertainty, resulting inblurry renderings that lack details. In this work, we propose NerfDiff, whichaddresses this issue by distilling the knowledge of a 3D-aware conditionaldiffusion model (CDM) into NeRF through synthesizing and refining a set ofvirtual views at test time. We further propose a novel NeRF-guided distillationalgorithm that simultaneously generates 3D consistent virtual views from theCDM samples, and finetunes the NeRF based on the improved virtual views. Ourapproach significantly outperforms existing NeRF-based and geometry-freeapproaches on challenging datasets, including ShapeNet, ABO, and Clevr3D.

NerfDiff: Single-image View Synthesis with NeRF-guided Distillation from 3D-aware Diffusion