U-GNN
U-GNN
Auto-generated stub. Edit this file to add more details.
A U-shaped encoder-decoder graph neural network architecture that employs a zero-padding pooling mechanism to enable multiscale feature extraction while maintaining original graph convolutionality.
Why It Matters
Provides a novel architectural approach to multiscale graph representation learning that avoids arbitrary graph coarsening via zero-padding.
Evidence
We introduce U-shaped encoder-decoder graph neural networks (U-GNNs) for stochastic graph signal generation using denoising diffusion processes.
Related Papers
Metadata & Links
- created_at
- 2026-04-24T05:51:55Z