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

Neural Radiance Field (NeRF) based rendering has attracted growing attentionthanks to its state-of-the-art (SOTA) rendering quality and wide applicationsin Augmented and Virtual Reality (AR/VR). However, immersive real-time (> 30FPS) NeRF based rendering enabled interactions are still limited due to the lowachievable throughput on AR/VR devices. To this end, we first profile SOTAefficient NeRF algorithms on commercial devices and identify two primary causesof the aforementioned inefficiency: (1) the uniform point sampling and (2) thedense accesses and computations of the required embeddings in NeRF.Furthermore, we propose RT-NeRF, which to the best of our knowledge is thefirst algorithm-hardware co-design acceleration of NeRF. Specifically, on thealgorithm level, RT-NeRF integrates an efficient rendering pipeline for largelyalleviating the inefficiency due to the commonly adopted uniform point samplingmethod in NeRF by directly computing the geometry of pre-existing points.Additionally, RT-NeRF leverages a coarse-grained view-dependent computingordering scheme for eliminating the (unnecessary) processing of invisiblepoints. On the hardware level, our proposed RT-NeRF accelerator (1) adopts ahybrid encoding scheme to adaptively switch between a bitmap- orcoordinate-based sparsity encoding format for NeRF's sparse embeddings, aimingto maximize the storage savings and thus reduce the required DRAM accesseswhile supporting efficient NeRF decoding; and (2) integrates both adual-purpose bi-direction adder & search tree and a high-density sparse searchunit to coordinate the two aforementioned encoding formats. Extensiveexperiments on eight datasets consistently validate the effectiveness ofRT-NeRF, achieving a large throughput improvement (e.g., 9.7x - 3,201x) whilemaintaining the rendering quality as compared with SOTA efficient NeRFsolutions.

RT-NeRF: Real-Time On-Device Neural Radiance Fields Towards Immersive AR/VR Rendering