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

Implicit neural representation has demonstrated promising results in viewsynthesis for large and complex scenes. However, existing approaches eitherfail to capture the fast-moving objects or need to build the scene graphwithout camera ego-motions, leading to low-quality synthesized views of thescene. We aim to jointly solve the view synthesis problem of large-scale urbanscenes and fast-moving vehicles, which is more practical and challenging. Tothis end, we first leverage a graph structure to learn the local scenerepresentations of dynamic objects and the background. Then, we design aprogressive scheme that dynamically allocates a new local scene graph trainedwith frames within a temporal window, allowing us to scale up therepresentation to an arbitrarily large scene. Besides, the training views ofurban scenes are relatively sparse, which leads to a significant decline inreconstruction accuracy for dynamic objects. Therefore, we design a frequencyauto-encoder network to encode the latent code and regularize the frequencyrange of objects, which can enhance the representation of dynamic objects andaddress the issue of sparse image inputs. Additionally, we employ lidar pointprojection to maintain geometry consistency in large-scale urban scenes.Experimental results demonstrate that our method achieves state-of-the-art viewsynthesis accuracy, object manipulation, and scene roaming ability. The codewill be open-sourced upon paper acceptance.

ProSGNeRF: Progressive Dynamic Neural Scene Graph with Frequency Modulated Auto-Encoder in Urban Scenes