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

We use neural radiance fields (NeRFs) to build interactive 3D environmentsfrom large-scale visual captures spanning buildings or even multiple cityblocks collected primarily from drones. In contrast to single object scenes (onwhich NeRFs are traditionally evaluated), our scale poses multiple challengesincluding (1) the need to model thousands of images with varying lightingconditions, each of which capture only a small subset of the scene, (2)prohibitively large model capacities that make it infeasible to train on asingle GPU, and (3) significant challenges for fast rendering that would enableinteractive fly-throughs.  To address these challenges, we begin by analyzing visibility statistics forlarge-scale scenes, motivating a sparse network structure where parameters arespecialized to different regions of the scene. We introduce a simple geometricclustering algorithm for data parallelism that partitions training images (orrather pixels) into different NeRF submodules that can be trained in parallel.  We evaluate our approach on existing datasets (Quad 6k and UrbanScene3D) aswell as against our own drone footage, improving training speed by 3x and PSNRby 12%. We also evaluate recent NeRF fast renderers on top of Mega-NeRF andintroduce a novel method that exploits temporal coherence. Our techniqueachieves a 40x speedup over conventional NeRF rendering while remaining within0.8 db in PSNR quality, exceeding the fidelity of existing fast renderers.

Mega-NeRF: Scalable Construction of Large-Scale NeRFs for Virtual Fly-Throughs