A conditional GAN with 3D sparse and dense convolutions reconstructs 4D radar tensor volumes from point clouds, with a percentile-based comparison showing 1% density balances compression and quality.
3drimr: 3d reconstruction and imaging via mmwave radar based on deep learning
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
4DR P2T: 4D Radar Tensor Synthesis with Point Clouds
A conditional GAN with 3D sparse and dense convolutions reconstructs 4D radar tensor volumes from point clouds, with a percentile-based comparison showing 1% density balances compression and quality.