Research map / Architectures and efficiency / RNNs and normalization
Localized, High-resolution Geographic Representations with Slepian Functions
Arjun Rao, Ruth Crasto, Tessa Ooms, David Rolnick, Konstantin Klemmer, Marc Rußwurm
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 |
|---|---|---|---|
| add global SH (Lg=10) to Slepian encoder (hybrid) (ablation) | Slepian encoder, Lr=120 · California Housing | R2: 0.69 -> 0.71 | Improved |
| add global SH to Slepian encoder (hybrid) (ablation) | Slepian encoder, Lr=120 · Japan Prefectures | accuracy: 0.89 -> 0.92 | Improved |
| add global SH to Slepian encoder (hybrid) (ablation) | Slepian encoder, Lr=120 · Arctic MSS | R2: 0.97 -> 0.98 | Improved |
| add global SH Lg=10 to Slepian-only encoder (ablation) | Slepian-only, LinNet · SINR IUCN | mAP: 0.011 -> 0.433 | Improved |
| add global SH Lg=10 to Slepian-only encoder (ablation) | Slepian-only, LinNet · SINR S&T | mAP: 0.264 -> 0.695 | Improved |
| increase SH degree from L=10 to L=40 | SH, LinNet · SINR IUCN | mAP: 0.342 -> 0.069 | Hurt |
| increase SH degree from L=10 to L=40 | SH, ResFCNet · SINR S&T | mAP: 0.736 -> 0.673 | Hurt |
| use Slepian (L=80) instead of SH (L=10) | SH L=10 · Arctic MSS | R2: 0.86 -> 0.97 | Improved |
| replace SH L=40 with Hybrid Slepian L=120 (1681 dims vs fewer) | SH (L=40) · Japan Prefectures | accuracy: 0.88 -> 0.92 | Improved |
| replace SH L=40 with Slepian L=40 | SH (L=40) · Japan Prefectures | accuracy: 0.88 -> 0.86 | Hurt |
| increase Slepian bandlimit from L=40 to L=120 (ablation) | Slepian · California Housing | R2: 0.64 -> 0.69 | Improved |
| use Hybrid Slepian instead of SH L=10 with LinNet | SH Lg=10 · SINR IUCN | mAP: 0.342 -> 0.478 | Improved |
Closest papers by meaning
- Geographic Location Encoding with Spherical Harmonics and Sinusoidal Representation Networks (ICLR 2024 · similarity 0.89)
- LocDiff: Identifying Locations on Earth by Diffusing in the Hilbert Space (NeurIPS 2025 · similarity 0.86)
- SatCLIP: Global, General-Purpose Location Embeddings with Satellite Imagery (AAAI 2025 · similarity 0.84)
- No Location Left Behind: Measuring and Improving the Fairness of Implicit Representations for Earth Data (ICLR 2025 · similarity 0.84)
- Poly2Vec: Polymorphic Fourier-Based Encoding of Geospatial Objects for GeoAI Applications (ICML 2025 · similarity 0.83)
- RANGE: Retrieval Augmented Neural Fields for Multi-Resolution Geo-Embeddings (CVPR 2025 · similarity 0.83)
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.
