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

In this work, we present a new multi-view depth estimation method thatutilizes both conventional reconstruction and learning-based priors over therecently proposed neural radiance fields (NeRF). Unlike existing neural networkbased optimization method that relies on estimated correspondences, our methoddirectly optimizes over implicit volumes, eliminating the challenging step ofmatching pixels in indoor scenes. The key to our approach is to utilize thelearning-based priors to guide the optimization process of NeRF. Our systemfirstly adapts a monocular depth network over the target scene by finetuning onits sparse SfM+MVS reconstruction from COLMAP. Then, we show that theshape-radiance ambiguity of NeRF still exists in indoor environments andpropose to address the issue by employing the adapted depth priors to monitorthe sampling process of volume rendering. Finally, a per-pixel confidence mapacquired by error computation on the rendered image can be used to furtherimprove the depth quality. Experiments show that our proposed frameworksignificantly outperforms state-of-the-art methods on indoor scenes, withsurprising findings presented on the effectiveness of correspondence-basedoptimization and NeRF-based optimization over the adapted depth priors. Inaddition, we show that the guided optimization scheme does not sacrifice theoriginal synthesis capability of neural radiance fields, improving therendering quality on both seen and novel views. Code is available athttps://github.com/weiyithu/NerfingMVS.

NerfingMVS: Guided Optimization of Neural Radiance Fields for Indoor Multi-view Stereo