OSGeo - Recent changes [en]

TorchGeo embeddings

Darkblueb — Mon, 15 Jun 2026 21:13:06 GMT

LLM-based Chat Assistant

Strk — Mon, 15 Jun 2026 17:34:12 GMT

TorchGeo embeddings

Darkblueb — Mon, 15 Jun 2026 16:40:12 GMT

Projects

Board Meeting 2026-03-31

Cvvergara — Mon, 15 Jun 2026 15:58:50 GMT

LDAP and OSGeo mantra

TorchGeo embeddings

Darkblueb — Mon, 15 Jun 2026 15:17:05 GMT

1. Foundation Models for Earth Observation (EO)

LLM-based Chat Assistant

Strk — Mon, 15 Jun 2026 14:53:27 GMT

TorchGeo embeddings

Darkblueb — Mon, 15 Jun 2026 14:29:22 GMT

2. Datasets

LLM-based Chat Assistant

Strk — Mon, 15 Jun 2026 13:04:20 GMT

TorchGeo embeddings

Darkblueb — Mon, 15 Jun 2026 03:01:57 GMT

Deatasets Auto-Edit =

Show changes

Talk:TorchGeo embeddings

Darkblueb — Sat, 13 Jun 2026 14:22:48 GMT

Created page with "'''Note''': AI LLM assistance was employed in the production of this document. {| class="wikitable" |+ LLM Engine contributions by operator |- ! Operator !! Tool !! M..."

New page

'''Note''': AI LLM assistance was employed in the production of this document.

{| class="wikitable"
|+ LLM Engine contributions by operator
|-
! Operator !! Tool !! Model 1 !! Model 2

|-
| '''darkblueb''' || ollama local-only || Qwen3.6-35B-A3B
|}

TorchGeo embeddings

Darkblueb — Sat, 13 Jun 2026 14:19:18 GMT

Key Papers

LLM-based Chat Assistant

Strk — Sat, 13 Jun 2026 11:04:16 GMT

@@ Line 9: / Line 9: @@
 }}
-arXiv:2601.13134v1 [cs.SE] 19 Jan 2026
+[[https://arxiv.org/html/2601.13134v1 arXiv:2601.13134v1]] [cs.SE] 19 Jan 2026
 '''Earth Embeddings as Products: Taxonomy, Ecosystem, and Standardized Access''' is a comprehensive survey that organizes existing geospatial embedding products into a structured taxonomy through a three-layer taxonomy:
@@ Line 61: / Line 61: @@
 A vision-language model designed to generate global location representations by aligning satellite imagery with geographic text embeddings, originally published in 2023 and widely studied.
-* '''MMEarth''' – [[https://arxiv.org/abs/2405.02771v2 [LINK]]] (ECCV 2024)
+* '''MMEarth''' – [[https://vishalned.github.io/mmearth/ SITE]] [[https://arxiv.org/abs/2405.02771v2 [arXiv]]] [[https://github.com/vishalned/MMEarth-data Datasets]] (ECCV 2024)
 A global dataset for multi-modal and geospatial representation learning. It consists of 12 modalities including pixel-level and image-level modalities from 1.2 million locations.
@@ Line 67: / Line 67: @@
 Transformers pre-trained via self-supervised objectives (e.g., DINO, MAE) are currently the state-of-the-art for extracting rich, transferable embeddings from remote sensing imagery.
-* '''DINO / DINOv2 / DINOv3''' – [[https://github.com/facebookresearch/dino [LINK]]] (Meta/INRIA, 2021–2023)
+* '''DINO / DINOv2 / DINOv3''' – [[https://ai.meta.com/research/dinov3/ [LINK]]] (Meta/INRIA, 2021–2023)
 Vision transformers trained with self-distillation without labels. DINOv2, in particular, produces highly discriminative embeddings that capture fine-grained geospatial features without downstream fine-tuning.
 * '''MAE (Masked Autoencoders)''' – [[https://arxiv.org/abs/2111.06377 [LINK]]] (CVPR 2021)

← Older revision		Revision as of 17:34, 15 June 2026
(One intermediate revision by the same user not shown)
Line 1:		Line 1:
−	~~# LLM-based Chat Assistant [rev:136081 2026-06-15T13:04:20Z]~~
−
−	~~# LLM-based Chat Assistant [rev:136049 2026-06-13T11:04:16Z]~~
−
−	~~# LLM-based Chat Assistant [rev:136031 2026-06-05T18:26:29Z]~~
−
	= LLM-based Chat Assistant =		= LLM-based Chat Assistant =

@@ Line 26: / Line 26: @@
 === Projects ===
 * '''Clay Foundation Model''' – [[https://huggingface.co/made-with-clay/Clay HuggingFace]] (2024)
 *: A multimodal foundation model for Earth using diverse data sources.
 * '''Earth Index Embeddings''' – [[https://www.earthgenome.org/earth-index EarthGenome]] (2025)
 *: A large-scale embedding system built from Earth observation data.
@@ Line 40: / Line 42: @@
 * '''Google Satellite Embedding (AlphaEarth)''' – [[https://developers.google.com/earth-engine/datasets/catalog/GOOGLE_SATELLITE_EMBEDDING_V1_ANNUAL LINK]] Google Earth Engine
 *: An early-stage embedding model using Google's global satellite data.
 === Key Papers ===

@@ Line 96: / Line 96: @@
 === LDAP and OSGeo mantra===
 * ACTION: Tim to write a motion in Loomio regarding keycloak
 === Actions ===

@@ Line 29: / Line 29: @@
 *: A multimodal foundation model for Earth using diverse data sources.
 * '''Major TOM''' – [[https://huggingface.co/Major-TOM MajorTOM]] AFrancis IGARSS 2024
-*: Expandable datasets and models for global EO coverage.
+*:  (Terrestrial Observation Metaset) is an open-source framework and massive satellite dataset created by the European Space Agency's Φ-lab.
 * '''Earth Index Embeddings''' – [[https://www.earthgenome.org/earth-index EarthGenome]] (2025)
 *: A large-scale embedding system built from Earth observation data.

@@ Line 1: / Line 1: @@
 # LLM-based Chat Assistant    [rev:136049 2026-06-13T11:04:16Z]
@@ Line 95: / Line 97: @@
 '''Main contact:''' ominiverdi, Lorenzo Becchi, or SAC channel.
-All the processes are currently run with "ominiverdi" user permissions.
+All the processes are currently run with "ominiverdi" user permissions, this should be changed ASAP.
 The starting process is invoked from directory /home/ominiverdi/github/opencode-chat-bridge using

@@ Line 49: / Line 49: @@
 == 2. Datasets ==
-Large-scale, open-access datasets play a central role in training and evaluating Earth foundation models.
+Large-scale, open-access datasets play a central role in training and evaluating Earth foundation models.  See the [[TorchGeo_embeddings#Deatasets_Auto-Edit|list]] below for a representative sample.
-== 3. Models & Methods ==
+== 3. Model Architectures for Embeddings ==
-These include both classical and cutting-edge machine learning approaches used in building Earth foundation models.
+The generation of robust Earth observation embeddings relies on specific model families pre-trained on large-scale geospatial data. The following architectures are central to the TorchGeo ecosystem and recent foundation model research, categorized by their primary learning paradigm.
-=== Core Methods ===
+=== Vision-Language Models ===
-* '''SatCLIP''' – [[https://arxiv.org/abs/2311.17179 LINK]] AAAI 2025 etc.
+These models align visual satellite imagery with textual or geographic metadata, creating unified embedding spaces for cross-modal retrieval.
-Vision-language model for global location representations.
-* '''MMEarth''' – [[https://arxiv.org/abs/2405.02771 LINK]] EU/CV 2024
+* '''SatCLIP''' – [[https://arxiv.org/abs/2311.17179v2 [LINK]]] (AAAI 2025)
-Multimodal pretext tasks for geospatial representation learning.
+A vision-language model designed to generate global location representations by aligning satellite imagery with geographic text embeddings, originally published in 2023 and widely studied.
-* '''ResNet''' – [[https://viso.ai/deep-learning/resnet-residual-neural-network/ LINK]] |KHe IEEE/CV 2016
-Baseline CNN architecture widely used in EO.
+* '''MMEarth''' – [[https://arxiv.org/abs/2405.02771v2 [LINK]]] (ECCV 2024)
-* '''ConvNeXt V2''' – [[https://github.com/facebookresearch/ConvNeXt-V2 LINK]] Woo et al., IEEE/CVF 2023
+A global dataset for multi-modal and geospatial representation learning. It consists of 12 modalities including pixel-level and image-level modalities from 1.2 million locations.
-Efficient ConvNet architecture using masked autoencoders (MAE).
-* '''DINO, DINOv2, DINOv3''' – [[https://dinov3.org/ LINK]] INRIA 2021–2023, META
+=== Self-Supervised Vision Transformers ===
-Vision transformers with self-supervised learning capabilities.
+Transformers pre-trained via self-supervised objectives (e.g., DINO, MAE) are currently the state-of-the-art for extracting rich, transferable embeddings from remote sensing imagery.
-* '''MAE (Masked Autoencoders)''' – [[https://arxiv.org/abs/2111.06377 LINK]] IEEE/CVF 2021
-Self-supervised learning for vision transformers.
+* '''DINO / DINOv2 / DINOv3''' – [[https://github.com/facebookresearch/dino [LINK]]] (Meta/INRIA, 2021–2023)
 === Distillation & Advanced Approaches ===

← Older revision		Revision as of 13:04, 15 June 2026
Line 1:		Line 1:
		+	# LLM-based Chat Assistant [rev:136049 2026-06-13T11:04:16Z]
		+
	# LLM-based Chat Assistant [rev:136031 2026-06-05T18:26:29Z]		# LLM-based Chat Assistant [rev:136031 2026-06-05T18:26:29Z]

Line 94:		Line 96:

	All the processes are currently run with "ominiverdi" user permissions.		All the processes are currently run with "ominiverdi" user permissions.
		+
		+	The starting process is invoked from directory /home/ominiverdi/github/opencode-chat-bridge using
		+
		+	<pre>
		+	bun connectors/matrix.ts
		+	</pre>

← Older revision		Revision as of 14:19, 13 June 2026
(5 intermediate revisions by the same user not shown)
Line 1:		Line 1:
		+	{{Infobox Paper
		+	\| title = Earth Embeddings as Products: Taxonomy, Ecosystem, and Standardized Access
		+	\| authors = Heng Fang<ref>†</ref>, Adam J Stewart<ref>‡</ref>, Isaac Corley<ref>§</ref>, Xiao Xiang Zhu<ref>*</ref>, Hossein Azizpour<ref>†</ref>
		+	\| affiliations =
		+	* † KTH Royal Institute of Technology, Stockholm, Sweden
		+	* ‡ Chair of Data Science in Earth Observation, Technical University of Munich, Munich, Germany
		+	* § Wherobots, San Antonio, USA
		+	\| arxiv = [https://arxiv.org/abs/2601.13134 2601.13134v1] (19 Jan 2026)
		+	}}
		+
		+	arXiv:2601.13134v1 [cs.SE] 19 Jan 2026
		+
		+	'''Earth Embeddings as Products: Taxonomy, Ecosystem, and Standardized Access''' is a comprehensive survey that organizes existing geospatial embedding products into a structured taxonomy through a three-layer taxonomy:
		+	Data, Tools, and Value. This research paper provides a detailed metadata atlas (resolution, license, etc.). It also proposes a unified integration by implementing standardized data loaders for these embeddings in [[https://www.osgeo.org/projects/torchgeo/ TorchGeo]] .
		+
		+	An overview landscape is proposed comprising:
		+	a) Analysis Frameworks & Tools
		+	b) Embeddings data artifacts
		+	c) Charting downstream application value, specifically mapping tasks and retrieval tasks.
		+
		+	Embeddings are differentiated as either location-typed, patch-typed, or pixel-typed. Details of existing products are shown. "We extend TorchGeo with a unified API that standardizes the loading and querying of diverse embedding products."
		+
		+	== 1. Foundation Models for Earth Observation (EO) ==
		+
		+	These are the leading projects that aim to build general-purpose models capable of representing Earth from satellite imagery and other geospatial modalities.
		+
		+	=== Projects ===
		+	* '''Clay Foundation Model''' – [[https://huggingface.co/made-with-clay/Clay HuggingFace]] (2024)
		+	*: A multimodal foundation model for Earth using diverse data sources.
		+	* '''Major TOM''' – [[https://huggingface.co/Major-TOM MajorTOM]] AFrancis IGARSS 2024
		+	*: Expandable datasets and models for global EO coverage.
		+	* '''Earth Index Embeddings''' – [[https://www.earthgenome.org/earth-index EarthGenome]] (2025)
		+	*: A large-scale embedding system built from Earth observation data.
		+	* '''Copernicus-Embed''' – [[https://github.com/zhu-xlab/Copernicus-FM LINK]] Zhu et al., AI4Copernicus Project
		+	*: Foundation model leveraging Copernicus Sentinel data.
		+	* '''Presto Embeddings''' – [[https://nasaharvest.github.io/presto-embeddings/ NASAHarvest]]
		+	*: Embedding framework for satellite time series and land use analysis.
		+	* '''Tessera Embeddings''' – [[https://geotessera.readthedocs.io/en/latest/ GeoTessera]] Docs / [[https://github.com/ucam-eo/tessera REPO]]
		+	*: Pixel-based Temporal spectral embeddings for Earth representation.
		+	* '''Google Satellite Embedding (AlphaEarth)''' – [[https://developers.google.com/earth-engine/datasets/catalog/GOOGLE_SATELLITE_EMBEDDING_V1_ANNUAL LINK]] Google Earth Engine
		+	*: An early-stage embedding model using Google's global satellite data.
		+	* '''OlmoEarth''' – [[https://allenai.org/olmoearth AllenAI]] (2025)
		+	*: Latent image modeling approach for multimodal Earth observation.
		+
		+	=== Key Papers ===
		+	* XXZhu 2025 [[https://www.nature.com/articles/s43247-025-03127-x LINK]] "On the Foundations of Earth Foundation Models" – ''Nature Computational Science''
		+	* CFBrown 2025 [[https://arxiv.org/abs/2507.22291 LINK]] "AlphaEarth Foundations"
		+	* KKlemmer 2023 [[https://arxiv.org/abs/2311.17179 LINK]] "SatCLIP: Global Location Embeddings with Satellite Imagery"
		+
		+	== 2. Datasets ==
		+
		+	Large-scale, open-access datasets play a central role in training and evaluating Earth foundation models.
		+
		+	=== Datasets ===
		+	* '''EuroSAT''' – [[https://zenodo.org/records/7711810\|Zenodo]]
		+	*: Land use classification dataset using Sentinel-2 satellite data.
		+	* '''EuroCrops''' – [[https://pmc.ncbi.nlm.nih.gov/articles/PMC10495462/\|PMC 10495462]]
		+	*: Crop type mapping dataset for Europe.
		+	* '''National Land Cover Database (NLCD)''' – [[https://www.ingentaconnect.com/content/asprs/pers/2004/00000070/00000007/art00005;jsessionid=2awgc0agiboi4.x-ic-live-02\|Photogrammetric Engineering & Remote Sensing]] (2001)
		+	*: USA land cover classes.
		+	* '''SSL4EO-S12''' – [[https://github.com/zhu-xlab/SSL4EO-S12\|IEEE Geoscience and Remote Sensing]] (2023)
		+	*: Multimodal, multitemporal dataset for self-supervised learning.
		+	* '''Copernicus-Pretrain''' – [[https://github.com/zhu-xlab/Copernicus-FM\|IEEE Geoscience and Remote Sensing]] (2023)
		+	*: An extension of the SSL4EO-S12 dataset to all major Sentinel missions (S1-S5P).
		+
		+	== 3. Models & Methods ==
		+
		+	These include both classical and cutting-edge machine learning approaches used in building Earth foundation models.
		+
		+	=== Core Methods ===
		+	* '''SatCLIP''' – [[https://arxiv.org/abs/2311.17179\|AAAI 2025]] etc.
		+	*: Vision-language model for global location representations.
		+	* '''MMEarth''' – [[https://arxiv.org/abs/2405.02771\|EU/CV 2024]]
		+	*: Multimodal pretext tasks for geospatial representation learning.
		+	* '''ResNet''' – [[https://viso.ai/deep-learning/resnet-residual-neural-network/\|KHe IEEE/CV 2016]]
		+	*: Baseline CNN architecture widely used in EO.
		+	* '''ConvNeXt V2''' – [[https://github.com/facebookresearch/ConvNeXt-V2\|Woo et al., IEEE/CVF 2023]]
		+	*: Efficient ConvNet architecture using masked autoencoders (MAE).
		+	* '''DINO, DINOv2, DINOv3''' – [[https://dinov3.org/\|INRIA 2021–2023, META]]
		+	*: Vision transformers with self-supervised learning capabilities.
		+	* '''MAE (Masked Autoencoders)''' – [[https://arxiv.org/abs/2111.06377\|IEEE/CVF 2021]]
		+	*: Self-supervised learning for vision transformers.
		+
		+	=== Distillation & Advanced Approaches ===
		+	* '''Distillation methods''' – Transfer knowledge from large models.
		+	* '''Neural plasticity-inspired models''' – [[https://arxiv.org/\|ZXiong, arXiv 2024]]: Inspired by biological learning mechanisms.
		+	* '''Multi-label guided soft contrastive learning''' – [[https://ieeexplore.ieee.org/\|YWang, IEEE TGRS, 2024]].
		+	* '''Barlow Twins''' – [[https://arxiv.org/\|Zbontar et al., arXiv 2021]]: Method for learning representations without contrastive loss.
		+	* '''Continual Barlow Twins''' – [[https://ieeexplore.ieee.org/\|IEEE JSTARS, 2023]]: Extends Barlow Twins to continual learning in EO segmentation.
		+
		+	== 4. Tools & Benchmarks ==
		+
		+	These are software systems and frameworks that support development, evaluation, or deployment of EO AI models.
		+
		+	=== Tools ===
		+	* '''TorchGeo''' – [[https://www.osgeo.org/projects/torchgeo/\|AJStewart ACM 2025]]
		+	*: PyTorch library for geospatial deep learning.
		+	* '''NeuCo-Bench''' – [[https://arxiv.org/html/2510.17914v1\|RVinge, arXiv 2025]]
		+	*: Benchmarking framework for neural embeddings in Earth observation.
		+	* '''GeoINRID''' – [[https://github.com/arjunarao619/GeoINRID\|GitHub: arjunarao619/GeoINRID]]
		+	*: Geospatial inference and representation learning toolkit.
		+
		+	=== Challenges ===
		+	* '''Embed2Scale Challenge''' – [[https://research.ibm.com/publications/the-2025-cvpr-earthvision-data-challenge-by-embed2scale\|CVPR CAlbrecht 2025]]
		+	*: Large-scale Earth vision challenge focused on scale-aware embeddings.
		+	* '''TerraMind Blue-Sky Challenge''' – [[https://arxiv.org/\|JJakubik, arXiv 2025]]
		+	*: Generative modeling for Earth observation.
		+
		+	== 5. Key Themes & Trends ==
		+
		+	# '''Foundation Models''': TorchGeo now includes data loaders designed for search/retrieval (Clay, Major TOM, Earth Index), and for dense prediction tasks like land cover mapping (Copernicus, Presto, Tessera, Google). TorchGeo allows us to enable fair, side-by-side benchmarking of different embedding models on the same downstream tasks, forming the basis for future experiments. Projects are encouraged to strengthen and improve explainability.
		+	#* '''Major TOM Notes''': Major TOM embeddings are not (yet) really product-oriented and are aimed with a similar purpose to the MT Core datasets - to make it easier to experiment and benchmark model outputs (hence, unlike TESSERA and AEF which came a few months after, MT embeddings do not have consistent or aggregated temporal scope). We haven't had enough time to finish off the preprint, but my current plan is to provide a simple MT Embedding benchmark at this year's EGU and integrate that into the arxiv pre-print. --Miko
		+	#* '''Earth Index / Earth Genome''': Use the Earth Index application (earthindex.ai) for non-technical users to use the embeddings we published on source.coop. Users of the web app (non-technical journalists, indigenous communities/allies, NGOs) have been our main focus. Users of the source.coop embeddings have generally been more technical folks interested in exploring/innovating in what's possible --BenStrong
		+	#* '''Clay''': Clay and Presto offer documented tutorials on generating new embeddings with their models. In CLAY, the encoder receives unmasked patches, latitude-longitude data, and timestep information. Notably, the last 2 embeddings from the encoder specifically represent the latitude-longitude and timestep embeddings.
		+
		+	# '''Self-Supervised Learning (SSL)''':
		+
		+	# '''Multimodal Integration''':
		+
		+	# '''Open Data & Tools''': Open-source projects (e.g., TorchGeo, Copernicus-Embed) and public datasets (EuroSAT, EuroCrops) are crucial for reproducibility and democratization of EO AI. Projects are encouraged to increase Input Data Diversity, and to adopt cloud-native data formats for geospatial data.
		+
		+	# '''Benchmarking''': Projects are encouraged to standardize in benchmarking. Benchmarks including NeuCo-Bench and Embed2Scale.
		+
		+	== Research Directions ==
		+
		+	* '''Unified Earth Foundation Models''':
		+	* '''Interpretability in EO AI''': Exploring how these embeddings can be interpreted by domain experts.
		+	* '''Ethics and Bias''': Investigating fairness and bias in global EO models trained on unevenly distributed data.
		+	* '''Edge Deployment''': Making these large foundation models deployable on resource-constrained platforms (e.g., for field use).
		+
		+	##-----------------------------------------------------
		+
	Notes on		Notes on
	'''EARTH EMBEDDINGS AS PRODUCTS: TAXONOMY, ECOSYSTEM, AND STANDARDIZED ACCESS'''		'''EARTH EMBEDDINGS AS PRODUCTS: TAXONOMY, ECOSYSTEM, AND STANDARDIZED ACCESS'''

@@ Line 1: / Line 1: @@
 = LLM-based Chat Assistant =
@@ Line 87: / Line 89: @@
 Language model: <code>opencode/big-pickle</code>, may change. Ask the bot for current model information.
-== Operations ==
+== Administration ==
 Check the bridge process on Gallery:
@@ Line 101: / Line 110: @@
 </pre>
 [[Category:Services]]
 [[Category:AI-Services]]