Srinidhi Hegde, Kaur Kullman, Thomas Grubb, Leslie Lait, Stephen Guimond, Matthias Zwicker
{"title":"NARVis: Neural Accelerated Rendering for Real-Time Scientific Point Cloud Visualization","authors":"Srinidhi Hegde, Kaur Kullman, Thomas Grubb, Leslie Lait, Stephen Guimond, Matthias Zwicker","doi":"arxiv-2407.19097","DOIUrl":null,"url":null,"abstract":"Exploring scientific datasets with billions of samples in real-time\nvisualization presents a challenge - balancing high-fidelity rendering with\nspeed. This work introduces a novel renderer - Neural Accelerated Renderer\n(NAR), that uses the neural deferred rendering framework to visualize\nlarge-scale scientific point cloud data. NAR augments a real-time point cloud\nrendering pipeline with high-quality neural post-processing, making the\napproach ideal for interactive visualization at scale. Specifically, we train a\nneural network to learn the point cloud geometry from a high-performance\nmulti-stream rasterizer and capture the desired postprocessing effects from a\nconventional high-quality renderer. We demonstrate the effectiveness of NAR by\nvisualizing complex multidimensional Lagrangian flow fields and photometric\nscans of a large terrain and compare the renderings against the\nstate-of-the-art high-quality renderers. Through extensive evaluation, we\ndemonstrate that NAR prioritizes speed and scalability while retaining high\nvisual fidelity. We achieve competitive frame rates of $>$ 126 fps for\ninteractive rendering of $>$ 350M points (i.e., an effective throughput of $>$\n44 billion points per second) using $\\sim$12 GB of memory on RTX 2080 Ti GPU.\nFurthermore, we show that NAR is generalizable across different point clouds\nwith similar visualization needs and the desired post-processing effects could\nbe obtained with substantial high quality even at lower resolutions of the\noriginal point cloud, further reducing the memory requirements.","PeriodicalId":501174,"journal":{"name":"arXiv - CS - Graphics","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2024-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - CS - Graphics","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2407.19097","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
Exploring scientific datasets with billions of samples in real-time
visualization presents a challenge - balancing high-fidelity rendering with
speed. This work introduces a novel renderer - Neural Accelerated Renderer
(NAR), that uses the neural deferred rendering framework to visualize
large-scale scientific point cloud data. NAR augments a real-time point cloud
rendering pipeline with high-quality neural post-processing, making the
approach ideal for interactive visualization at scale. Specifically, we train a
neural network to learn the point cloud geometry from a high-performance
multi-stream rasterizer and capture the desired postprocessing effects from a
conventional high-quality renderer. We demonstrate the effectiveness of NAR by
visualizing complex multidimensional Lagrangian flow fields and photometric
scans of a large terrain and compare the renderings against the
state-of-the-art high-quality renderers. Through extensive evaluation, we
demonstrate that NAR prioritizes speed and scalability while retaining high
visual fidelity. We achieve competitive frame rates of $>$ 126 fps for
interactive rendering of $>$ 350M points (i.e., an effective throughput of $>$
44 billion points per second) using $\sim$12 GB of memory on RTX 2080 Ti GPU.
Furthermore, we show that NAR is generalizable across different point clouds
with similar visualization needs and the desired post-processing effects could
be obtained with substantial high quality even at lower resolutions of the
original point cloud, further reducing the memory requirements.