Research map / Graph learning / GNN architectures
Fourier-Based and Rational Graph Filters for Spectral Processing
Giuseppe Patanè
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use diffusion functions (48 functions) instead of 60 Laplacian eigenfunctions as functional-map basis | functional maps · SHREC'10 non-isometric shapes | mean correspondence error | Improved |
| approximate the spectral kernel with a rational polynomial instead of truncated eigen-expansion | spectral wavelets · small-scale diffusion wavelet | approximation accuracy and efficiency | Improved |
| use harmonic or diffusion kernels instead of Laplacian eigenfunctions for reconstruction | signal reconstruction · 3D shape geometry | l-infinity approximation error | No clear effect |
| add more diffusion functions at a scale (and use small scales) (ablation) | signal smoothing · noisy 3D shape coordinates | relative l-infinity approximation error | Improved |
| use rational spectral wavelet exp(-st)s on partial shapes with different connectivity instead of partial functional maps | partial shape correspondence · FARM partial dataset | mean geodesic distance | Improved |
| use rational spectral wavelet instead of the best partial functional maps method on similar connectivity | partial shape correspondence · SHREC'16 partial cut | mean geodesic distance | No clear effect |
Closest papers by meaning
- A General Graph Spectral Wavelet Convolution via Chebyshev Order Decomposition (ICML 2025 · similarity 0.90)
- How Universal Polynomial Bases Enhance Spectral Graph Neural Networks: Heterophily, Over-smoothing, and Over-squashing (ICML 2024 · similarity 0.88)
- Position: Spectral GNNs Rely Less on Graph Fourier Basis than Conceived (ICML 2025 · similarity 0.87)
- WaveNet: Tackling Non-stationary Graph Signals via Graph Spectral Wavelets (AAAI 2024 · similarity 0.87)
- HoloNets: Spectral Convolutions do extend to Directed Graphs (ICLR 2024 · similarity 0.87)
- GrokFormer: Graph Fourier Kolmogorov-Arnold Transformers (ICML 2025 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
