UniverSat is a ViT-style model with a universal patch encoder enabling self-supervised training on heterogeneous multimodal Earth observation data from varying resolutions and sensors.
TESSERA: Temporal Embeddings of Surface Spectra for Earth Representation and Analysis
11 Pith papers cite this work, alongside 37 external citations. Polarity classification is still indexing.
11
Pith papers citing it
37
external citations · Pith
abstract
Satellite Earth-observation (EO) time series in the optical and microwave ranges of the electromagnetic spectrum are often irregular due to orbital patterns and cloud obstruction. Compositing addresses these issues but loses information with respect to vegetation phenology, which is critical for many downstream tasks. Instead, we present TESSERA, a pixel-wise foundation model for multi-modal (Sentinel-1/2) EO time series that learns robust, label-efficient embeddings. During model training, TESSERA uses Barlow Twins and sparse random temporal sampling to enforce invariance to the selection of valid observations. We employ two key regularizers: global shuffling to decorrelate spatial neighborhoods and mix-based regulation to improve invariance under extreme sparsity. We find that for diverse classification, segmentation, and regression tasks, TESSERA embeddings deliver state-of-the-art accuracy with high label efficiency, often requiring only a small task head and minimal computation. To democratize access, adhere to FAIR - principles, and simplify use, we release global, annual, 10m, pixel-wise int8 embeddings together with open weights/code and lightweight adaptation heads, thus providing practical tooling for large-scale retrieval and inference at planetary scale. All code and data are available at: https://github.com/ucam-eo/tessera.
citation-role summary
background 2
citation-polarity summary
years
2026 11verdicts
UNVERDICTED 11roles
background 2polarities
background 2representative citing papers
Biomazon is a multimodal benchmark dataset pairing GEDI RH profiles and AGBD targets with Sentinel, ALOS, DEM, and other predictors for joint 3D forest structure and biomass modeling in the Amazon.
Fusing embeddings from four Earth models (AlphaEarth, Tessera, GeoCLIP, SatCLIP) outperforms the best single model on four of six tasks, with gains depending on task and location.
DeluluNet enables continued prediction under modality substitution, addition, or subsets by training a multi-modal model from a unimodal teacher and unlabeled multimodal data via modality hallucination.
CityRep is a new multi-city, multi-task benchmark with spatial splits for evaluating urban representation embeddings across modalities and locations.
FLUXtrapolation is a benchmark for domain generalization in ecosystem flux upscaling using temporal, spatial, and temperature-based extrapolation scenarios, with pilot results showing model separation on tail and multi-scale metrics.
SSDM decouples global geospatial embeddings into structural modulation and semantic injection pathways to improve accuracy and consistency in high-resolution remote sensing land cover mapping.
LIANet encodes multi-temporal Earth observation data into a coordinate-based neural field that supports label-only fine-tuning for downstream tasks without access to raw imagery.
Releases a publicly available, collocated multi-sensor dataset of Landsat, Sentinel-1, GOES-R and microwave observations for urban heat studies across 48 cities.
Position paper identifies structural challenges in applying generic agentic AI to Earth Observation and outlines design principles for EO-native agents focused on geospatial state and validity.
Perspective paper calling for unified spatial representation learning that integrates raster imagery with vector semantics in geospatial foundation models.
citing papers explorer
-
UniverSat: Resolution- and Modality-Agnostic Transformers for Earth Observation
UniverSat is a ViT-style model with a universal patch encoder enabling self-supervised training on heterogeneous multimodal Earth observation data from varying resolutions and sensors.
-
Biomazon: A Multimodal Dataset for 3D Forest Structure and Biomass Modeling in the Amazon Basin
Biomazon is a multimodal benchmark dataset pairing GEDI RH profiles and AGBD targets with Sentinel, ALOS, DEM, and other predictors for joint 3D forest structure and biomass modeling in the Amazon.
-
Better Together: Evaluating the Complementarity of Earth Embedding Models
Fusing embeddings from four Earth models (AlphaEarth, Tessera, GeoCLIP, SatCLIP) outperforms the best single model on four of six tasks, with gains depending on task and location.
-
Changing Modalities: Adapting Remote Sensing Models to New Satellites and Sensors
DeluluNet enables continued prediction under modality substitution, addition, or subsets by training a multi-modal model from a unimodal teacher and unlabeled multimodal data via modality hallucination.
-
CITYREP: A Unified Benchmark for Urban Representations Across Cities, Tasks, and Modalities
CityRep is a new multi-city, multi-task benchmark with spatial splits for evaluating urban representation embeddings across modalities and locations.
-
FLUXtrapolation: A benchmark on extrapolating ecosystem fluxes
FLUXtrapolation is a benchmark for domain generalization in ecosystem flux upscaling using temporal, spatial, and temperature-based extrapolation scenarios, with pilot results showing model separation on tail and multi-scale metrics.
-
Structure-Semantic Decoupled Modulation of Global Geospatial Embeddings for High-Resolution Remote Sensing Mapping
SSDM decouples global geospatial embeddings into structural modulation and semantic injection pathways to improve accuracy and consistency in high-resolution remote sensing land cover mapping.
-
Location Is All You Need: Continuous Spatiotemporal Neural Representations of Earth Observation Data
LIANet encodes multi-temporal Earth observation data into a coordinate-based neural field that supports label-only fine-tuning for downstream tasks without access to raw imagery.
-
Urban Heat MiniCubes: An AI-Ready dataset for urban heat research
Releases a publicly available, collocated multi-sensor dataset of Landsat, Sentinel-1, GOES-R and microwave observations for urban heat studies across 48 cities.
-
Agentic AI for Remote Sensing: Technical Challenges and Research Directions
Position paper identifies structural challenges in applying generic agentic AI to Earth Observation and outlines design principles for EO-native agents focused on geospatial state and validity.
-
Spatial Representation Learning Beyond Pixels: Unifying Raster Data and Vector Semantics for Human-Centric Geospatial Foundation Models
Perspective paper calling for unified spatial representation learning that integrates raster imagery with vector semantics in geospatial foundation models.