Landsat program
PulseAugur coverage of Landsat program — every cluster mentioning Landsat program across labs, papers, and developer communities, ranked by signal.
3 day(s) with sentiment data
-
New benchmark VHR-Bench evaluates foundation models on high-res Earth imagery
Researchers have introduced VHR-Bench, a new benchmark suite designed to evaluate foundation models on very high-resolution (VHR) Earth observation imagery. The study found that models pretrained on large natural image …
-
New EarthRecord benchmark tests AI agents on four decades of satellite imagery
Researchers have introduced EarthRecord, a new benchmark designed to evaluate multimodal AI agents' ability to analyze land-change over four decades using raw satellite imagery. Unlike previous benchmarks, EarthRecord u…
-
GeoAI model predicts wildfire structure loss with spatial validation
A new research paper details a GeoAI workflow designed to predict urban wildfire structure loss, using data from the January 2025 Palisades Fire. The study integrated building information, vegetation indices, terrain da…
-
Hugging Face releases OlmoEarth v1.3 satellite embedding model
Hugging Face has released OlmoEarth v1.3, a new state-of-the-art satellite embedding model. This release includes precomputed annual embeddings for the entire planet at 10-meter resolution, available in 64- and 128-dime…
-
Deep Learning Maps Indian Forest Plantations Using Satellite Data
Researchers have developed a deep learning approach to map small forest plantations in India's Godavari Districts. Utilizing satellite data from the Harmonized Landsat Sentinel-2 S10 product, the study employed a pixel-…
-
New deep-learning framework improves carbon flux prediction accuracy
Researchers have developed a deep-learning framework called Footprint-Aware Regression (FAR) to improve the accuracy of carbon flux predictions. This new approach accounts for the spatial footprint of eddy-covariance (E…
-
FORMSpoT maps forest disturbances at 1.5m resolution across France
Researchers have developed FORMSpoT, a novel system for monitoring forest disturbances with unprecedented 1.5 m resolution across France. This system utilizes satellite imagery from SPOT-6/7 and a transformer model (PVT…
-
New AI method uses satellite data for poverty mapping with uncertainty awareness
Researchers have developed a new machine learning method using satellite imagery to estimate poverty levels in Africa, aiming to provide more reliable data for public policy. This uncertainty-aware approach, based on si…
-
New benchmark dataset HeatCast enables neighborhood-scale LST forecasting
Researchers have introduced HeatCast, a new benchmark dataset for forecasting land surface temperature (LST) at a neighborhood scale across 124 U.S. cities. This dataset, derived from Landsat imagery, provides monthly L…
-
New SLED method offers scalable, cost-effective geospatial data encoding
Researchers have developed a new method called Scalable Location Encoding via Distillation (SLED) for creating efficient location encoders from geospatial data. Unlike previous methods that rely on computationally expen…
-
Google retracts AI image generator from Google Earth amid disinformation fears
Google has introduced and subsequently rolled back an AI feature in Google Earth that allowed users to generate fabricated satellite images based on text prompts. This tool, powered by Google's Nano Banana model, raised…
-
New GAN reconstructs satellite land surface temperature data with high accuracy
Researchers have developed a novel Multimodal Fast Fourier Convolutional GAN designed to reconstruct land surface temperature (LST) data from satellite imagery, specifically addressing gaps caused by cloud cover. This m…
-
AI predicts poverty using satellite images and LLM-generated text
Researchers have developed a multimodal framework to predict household wealth in African neighborhoods using satellite imagery and text generated by AI. The framework combines vision models with LLM-generated text and w…
-
GlacierCastAI uses satellite imagery and climate data to predict glacier retreat
Researchers have developed GlacierCastAI, a novel deep learning model designed to predict glacier retreat using a combination of multi-modal satellite imagery and climate data. The model integrates data from the Landsat…
-
HLS-GPT Transformer reconstructs NASA satellite reflectance data
Researchers have developed HLS-GPT, a large-scale generative pretrained Transformer model designed to reconstruct NASA's Harmonized Landsat and Sentinel-2 (HLS) surface reflectance data. This model utilizes a hierarchic…
-
Vision transformer maps 38 years of US forest disturbances
Researchers have developed a deep learning framework using a vision transformer to map forest disturbances across the contiguous United States over a 38-year period. This approach simultaneously models temporal trajecto…
-
Hybrid ML model improves forest height estimation using satellite data
Researchers have developed a hybrid machine learning model that improves forest height estimation by integrating data from TanDEM-X and Landsat satellites. This enhanced model incorporates optical Landsat data to provid…
-
New EO-Gym environment trains AI agents for interactive Earth Observation analysis
Researchers have introduced EO-Gym, an interactive framework designed for Earth Observation (EO) agents. This environment supports multimodal analysis and tool usage, simulating real-world EO tasks that often involve ex…
-
Machine learning model maps soil salinity in Bangladesh
Researchers have developed a machine-learning framework to map and predict soil salinity in Satkhira, Bangladesh, using field data and satellite imagery. An Extreme Gradient Boosting model, trained on 205 soil samples, …
-
Satellite foundation model Tempov maps wealth and poverty in Africa
Researchers have developed Tempov, a novel satellite foundation model designed to enhance wealth monitoring in low- and middle-income countries. This model, trained using self-supervision on millions of satellite image …