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 paper, we introduce a deep multi-view stereo (MVS) system thatjointly predicts depths, surface normals and per-view confidence maps. The keyto our approach is a novel solver that iteratively solves for per-view depthmap and normal map by optimizing an energy potential based on the locallyplanar assumption. Specifically, the algorithm updates depth map by propagatingfrom neighboring pixels with slanted planes, and updates normal map with localprobabilistic plane fitting. Both two steps are monitored by a customizedconfidence map. This solver is not only effective as a post-processing tool forplane-based depth refinement and completion, but also differentiable such thatit can be efficiently integrated into deep learning pipelines. Our multi-viewstereo system employs multiple optimization steps of the solver over theinitial prediction of depths and surface normals. The whole system can betrained end-to-end, decoupling the challenging problem of matching pixelswithin poorly textured regions from the cost-volume based neural network.Experimental results on ScanNet and RGB-D Scenes V2 demonstratestate-of-the-art performance of the proposed deep MVS system on multi-viewdepth estimation, with our proposed solver consistently improving the depthquality over both conventional and deep learning based MVS pipelines. Code isavailable at https://github.com/thuzhaowang/idn-solver.

A Confidence-based Iterative Solver of Depths and Surface Normals for Deep Multi-view Stereo