mlsanityData to 7 Oct 2026

TorchSpatial: A Location Encoding Framework and Benchmark for Spatial Representation Learning

arXiv 2406.15658 ↗First posted 21 Jun 2024cs.CV cs.AIcited by 29

Not on the map. Why?

Abstract (arXiv)

Spatial representation learning (SRL) aims at learning general-purpose neural network representations from various types of spatial data (e.g., points, polylines, polygons, networks, images, etc.) in their native formats. Learning good spatial representations is a fundamental problem for various downstream applications such as species distribution modeling, weather forecasting, trajectory generation, geographic question answering, etc. Even though SRL has become the foundation of almost all geospatial artificial intelligence (GeoAI) research, we have not yet seen significant efforts to develop an extensive deep learning framework and benchmark to support SRL model development and evaluation. To fill this gap, we propose TorchSpatial, a learning framework and benchmark for location (point) encoding, which is one of the most fundamental data types of spatial representation learning. TorchSpatial contains three key components: 1) a unified location encoding framework that consolidates 15 commonly recognized location encoders, ensuring scalability and reproducibility of the implementations; 2) the LocBench benchmark tasks encompassing 7 geo-aware image classification and 10 geo-aware image regression datasets; 3) a comprehensive suite of evaluation metrics to quantify geo-aware model's overall performance as well as their geographic bias, with a novel Geo-Bias Score metric. Finally, we provide a detailed analysis and insights into the model performance and geographic bias of different location encoders. We believe TorchSpatial will foster future advancement of spatial representation learning and spatial fairness in GeoAI research. The TorchSpatial model framework and LocBench benchmark are available at https://github.com/seai-lab/TorchSpatial, and the Geo-Bias Score evaluation framework is available at https://github.com/seai-lab/PyGBS.

Key facts. Cited by 29 papers in the map (data through 7 Oct 2026). Builds on 48 of its 100 references that are in the map. First posted on arXiv 21 Jun 2024.

Builds on

48 of 100 references.

References · 48

  1. 2021On the Opportunities and Risks of Foundation Modelscited by 3,961
  2. 2019Optuna: A Next-generation Hyperparameter Optimization Frameworkcited by 2,455
  3. 2019Fast Graph Representation Learning with PyTorch Geometriccited by 2,414
  4. 2017VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detectioncited by 1,323
  5. 2022SatMAE: Pre-training Transformers for Temporal and Multi-Spectral Satellite Imagerycited by 402
  6. 2017Functional Map of the Worldcited by 338
  7. 2017A Neural Representation of Sketch Drawingscited by 311
  8. 2023VoxelNeXt: Fully Sparse VoxelNet for 3D Object Detection and Trackingcited by 221
  9. 2020Geography-Aware Self-Supervised Learningcited by 190
  10. 2023GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationcited by 185
  11. 2023SatCLIP: Global, General-Purpose Location Embeddings with Satellite Imagerycited by 148
  12. 2023GEO-Bench: Toward Foundation Models for Earth Monitoringcited by 122
36 more
  1. 2018Efficient Interactive Annotation of Segmentation Datasets with Polygon-RNN++cited by 119
  2. 2019Presence-Only Geographical Priors for Fine-Grained Image Classificationcited by 109
  3. 2019PolyTransform: Deep Polygon Transformer for Instance Segmentationcited by 92
  4. 2024Large Language Models are Geographically Biasedcited by 91
  5. 2017Annotating Object Instances with a Polygon-RNNcited by 88
  6. 2020Multi-Scale Representation Learning for Spatial Feature Distributions using Grid Cellscited by 88
  7. 2020A Generalizable and Accessible Approach to Machine Learning with Global Satellite Imagerycited by 87
  8. 2021TorchAudio: Building Blocks for Audio and Speech Processingcited by 83
  9. 2021TorchGeo: Deep Learning With Geospatial Datacited by 73
  10. 2023CSP: Self-Supervised Contrastive Spatial Pre-Training for Geospatial-Visual Representationscited by 67
  11. 2023GeoLLM: Extracting Geospatial Knowledge from Large Language Modelscited by 62
  12. 2023Geographic Location Encoding with Spherical Harmonics and Sinusoidal Representation Networkscited by 61
  13. 2020LSTM-TrajGAN: A Deep Learning Approach to Trajectory Privacy Protectioncited by 57
  14. 2023Spatial Implicit Neural Representations for Global-Scale Species Mappingcited by 57
  15. 2019Reconstructing continuous distributions of 3D protein structure from cryo-EM imagescited by 56
  16. 2021A Review of Location Encoding for GeoAI: Methods and Applicationscited by 53
  17. 2021SustainBench: Benchmarks for Monitoring the Sustainable Development Goals with Machine Learningcited by 53
  18. 2022Connecting the Dots: Floorplan Reconstruction Using Two-Level Queriescited by 48
  19. 2019Geo-Aware Networks for Fine-Grained Recognitioncited by 44
  20. 2022EarthNets: Empowering AI in Earth Observationcited by 40
  21. 2023Sphere2Vec: A General-Purpose Location Representation Learning over a Spherical Surface for Large-Scale Geospatial Predictionscited by 38
  22. 2015Improving Image Classification with Location Contextcited by 33
  23. 2024Img2Loc: Revisiting Image Geolocalization using Multi-modality Foundation Models and Image-based Retrieval-Augmented Generationcited by 31
  24. 2022Geographic and Geopolitical Biases of Language Modelscited by 29
  25. 2021Geographic Question Answering: Challenges, Uniqueness, Classification, and Future Directionscited by 24
  26. 2020SE-KGE: A Location-Aware Knowledge Graph Embedding Model for Geographic Question Answering and Spatial Semantic Liftingcited by 23
  27. 2021Toward Foundation Models for Earth Monitoring: Proposal for a Climate Change Benchmarkcited by 22
  28. 2022Towards General-Purpose Representation Learning of Polygonal Geometriescited by 21
  29. 2023City Foundation Models for Learning General Purpose Representations from OpenStreetMapcited by 21
  30. 2018Learning Grid Cells as Vector Representation of Self-Position Coupled with Matrix Representation of Self-Motioncited by 18
  31. 2019Transferring Multiscale Map Styles Using Generative Adversarial Networkscited by 11
  32. 2018Deep Learning for Classification Tasks on Geospatial Vector Polygonscited by 9
  33. 2020Joint Spatio-Textual Reasoning for Answering Tourism Questionscited by 9
  34. 2023CATS: Conditional Adversarial Trajectory Synthesis for Privacy-Preserving Trajectory Data Publication Using Deep Learning Approachescited by 6
  35. 2019DDSL: Deep Differentiable Simplex Layer for Learning Geometric Signalscited by 5
  36. 2024Probing the Information Theoretical Roots of Spatial Dependence Measurescited by 3

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 4; 2025: 15; 2026: 10.

29 papers.

Sort:

By area:

  1. 2025Spatial-RAG: Spatial Retrieval Augmented Generation for Real-World Geospatial Reasoning Questionscited by 13
  2. 2025A Survey of Large Language Model-Powered Spatial Intelligence Across Scales: Advances in Embodied Agents, Smart Cities, and Earth Sciencecited by 11
  3. 2025Measuring the Intrinsic Dimension of Earth Representationscited by 10
  4. 2024REO-VLM: Transforming VLM to Meet Regression Challenges in Earth Observationcited by 9
  5. 2025GAIR: Location-Aware Self-Supervised Contrastive Pre-Training with Geo-Aligned Implicit Representationscited by 9
  6. 2025LocDiff: Identifying Locations on Earth by Diffusing in the Hilbert Spacecited by 6
  7. 2025Whose Truth? Pluralistic Geo-Alignment for (Agentic) AIcited by 6
  8. 2024Generating Binary Species Range Mapscited by 4
  9. 2025S2Vec: Self-Supervised Geospatial Embeddings for the Built Environmentcited by 4
  10. 2025MoRA: Mobility as the Backbone for Geospatial Representation Learning at Scalecited by 4
  11. 2026Localized, High-resolution Geographic Representations with Slepian Functionscited by 3
  12. 2026CITYREP: A Unified Benchmark for Urban Representations Across Cities, Tasks, and Modalitiescited by 3
17 more
  1. 2024GeoConformal prediction: a model-agnostic framework of measuring the uncertainty of spatial predictioncited by 2
  2. 2025GeoBS: Information-Theoretic Quantification of Geographic Bias in AI Modelscited by 2
  3. 2025OBSR: Open Benchmark for Spatial Representationscited by 2
  4. 2025Latent Domain Modeling Improves Robustness to Geographic Shiftscited by 1
  5. 2025ZooplanktonBench: A Geo-Aware Zooplankton Recognition and Classification Dataset from Marine Observationscited by 1
  6. 2025Towards Unified Vision Language Models for Forest Ecological Analysis in Earth Observationcited by 1
  7. 2026A Proxy Consistency Loss for Grounded Fusion of Earth Observation and Location Encoderscited by 1
  8. 2026SLED: Scalable Location Encoding via Distillationcited by 1
  9. 2026CoST: Semantic-Aware Urban Understanding via Spatial-Temporal Alignmentcited by 1
  10. 2024A Unified Framework for Next-Gen Urban Forecasting via LLM-driven Dependency Retrieval and GeoTransformercited by 0
  11. 2025Efficient Mixture of Geographical Species for On Device Wildlife Monitoringcited by 0
  12. 2025Geospatial Foundational Embedder: Top-1 Winning Solution on EarthVision Embed2Scale Challenge (CVPR 2025)cited by 0
  13. 2026Geography According to ChatGPT -- How Generative AI Represents and Reasons about Geographycited by 0
  14. 2026Geographic Bias and Diversity in AI Evaluationcited by 0
  15. 2026Textual Supervision Enhances Geospatial Representations in Vision-Language Modelscited by 0
  16. 2026Do Location Encoders Capture Spatial Effects? A GeoShapley Benchmark Across Scalescited by 0
  17. 2026MABLE: Masked Autoencoding with Bi-Lipschitz Decoding for Embeddings and Graph Metric Learningcited by 0