Make the Fastest Faster: Importance Mask Synthesis for Interactive Volume Visualization using Reconstruction Neural Networks
\authororcid Jianxin Sun0000-0002-9627-9397, \authororcid David Lenz0000-0002-2587-2783, \authororcid Hongfeng Yu0000-0002-0596-8227, and \authororcid Tom Peterka0000-0002-0525-3205 · 2025
Read Make the Fastest Faster: Importance Mask Synthesis for Interactive Volume Visualization using Reconstruction Neural Networks by \authororcid Jianxin Sun0000-0002-9627-9397, \authororcid David Lenz0000-0002-2587-2783, \authororcid Hongfeng Yu0000-0002-0596-8227, and \authororcid Tom Peterka0000-0002-0525-3205 (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-06053.txt.
Description
Jianxin Sun, David Lenz, Hongfeng Yu, and Tom Peterka speed up the viewing of huge 3D data sets by predicting which pixels matter before rendering them. Their 2025 paper builds on a recent trick: render only some pixels the expensive way and let a neural network paint in the rest. Existing versions still render too much, either on a fixed pattern or on one guessed from a rough preview.
Their networks learn the importance mask directly from the viewer's position and settings, jointly tuned with the data set and with the reconstruction network that will consume the result. Flexible layers let the same framework serve different inpainting styles. In tests on scientific volumes, the method cut overall latency further at no cost in quality, and it could even fine-tune a pretrained reconstruction network in place. The authors frame it as making the fastest method faster by rendering less, more cleverly.
Catalog record
| CREATOR | \authororcid Jianxin Sun0000-0002-9627-9397, \authororcid David Lenz0000-0002-2587-2783, \authororcid Hongfeng Yu0000-0002-0596-8227, and \authororcid Tom Peterka0000-0002-0525-3205 |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 68 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2502-06053