DM3D introduces offset-guided differentiable scanning and continuity-aware state updates in a Mamba-based model to achieve state-of-the-art or competitive performance on point cloud classification, few-shot learning, and part segmentation.
Combining recurrent, convolutional, and continuous-time models with linear state space layers
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
DM3D: Deformable Mamba via Offset-Guided Differentiable Scanning for Point Cloud Understanding
DM3D introduces offset-guided differentiable scanning and continuity-aware state updates in a Mamba-based model to achieve state-of-the-art or competitive performance on point cloud classification, few-shot learning, and part segmentation.