RadTwin conditions a neural radio-propagation model on scene point clouds via physics-informed sparse attention, achieving 0.846 SSIM and 0.023 LPIPS on dynamic indoor scenes without retraining.
Deep point cloud reconstruction
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.NI 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
RadTwin: Generalizable Wireless Digital Twin for Dynamic Environments
RadTwin conditions a neural radio-propagation model on scene point clouds via physics-informed sparse attention, achieving 0.846 SSIM and 0.023 LPIPS on dynamic indoor scenes without retraining.