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/

Zendy Plus

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

Zendy Tools

Zendy Open

Novel view synthesis via Neural Radiance Fields (NeRFs) or 3D GaussianSplatting (3DGS) typically necessitates dense observations with hundreds ofinput images to circumvent artifacts. We introduce Deceptive-NeRF/3DGS toenhance sparse-view reconstruction with only a limited set of input images, byleveraging a diffusion model pre-trained from multiview datasets. Differentfrom using diffusion priors to regularize representation optimization, ourmethod directly uses diffusion-generated images to train NeRF/3DGS as if theywere real input views. Specifically, we propose a deceptive diffusion modelturning noisy images rendered from few-view reconstructions into high-qualityphotorealistic pseudo-observations. To resolve consistency amongpseudo-observations and real input views, we develop an uncertainty measure toguide the diffusion model's generation. Our system progressively incorporatesdiffusion-generated pseudo-observations into the training image sets,ultimately densifying the sparse input observations by 5 to 10 times. Extensiveexperiments across diverse and challenging datasets validate that our approachoutperforms existing state-of-the-art methods and is capable of synthesizingnovel views with super-resolution in the few-view setting.

Deceptive-NeRF/3DGS: Diffusion-Generated Pseudo-Observations for High-Quality Sparse-View Reconstruction