Landsat-Bench: Datasets and Benchmarks for Landsat Foundation Models

• URL: http://arxiv.org/abs/2506.08780v1
• Date: Tue, 10 Jun 2025 13:24:55 GMT
• Title: Landsat-Bench: Datasets and Benchmarks for Landsat Foundation Models
• Authors: Isaac Corley, Lakshay Sharma, Ruth Crasto,
• Abstract summary: We introduce Landsat-Bench, a suite of three benchmarks with Landsat imagery that adapt from existing remote sensing datasets.<n>We establish baseline and standardized evaluation methods across both common architectures and Landsat foundation models pretrained on the SSL4EO-L dataset.
• Score: 0.6144680854063939
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The Landsat program offers over 50 years of globally consistent Earth imagery. However, the lack of benchmarks for this data constrains progress towards Landsat-based Geospatial Foundation Models (GFM). In this paper, we introduce Landsat-Bench, a suite of three benchmarks with Landsat imagery that adapt from existing remote sensing datasets -- EuroSAT-L, BigEarthNet-L, and LC100-L. We establish baseline and standardized evaluation methods across both common architectures and Landsat foundation models pretrained on the SSL4EO-L dataset. Notably, we provide evidence that SSL4EO-L pretrained GFMs extract better representations for downstream tasks in comparison to ImageNet, including performance gains of +4% OA and +5.1% mAP on EuroSAT-L and BigEarthNet-L.

Related papers

• GeoLocSFT: Efficient Visual Geolocation via Supervised Fine-Tuning of Multimodal Foundation Models [4.956977275061966]
  GeoLocSFT is trained with only 2700 carefully selected image-GPS pairs from our geographically diverse MR600k dataset.<n>Despite this limited data, our SFT-centric approach substantially improves over baseline models.<n>Our findings highlight the power of high-quality supervision and efficient SFT for planet-scale image geolocation.
  arXiv  Detail & Related papers  (2025-06-02T03:16:19Z)
• OpenEarthMap-SAR: A Benchmark Synthetic Aperture Radar Dataset for Global High-Resolution Land Cover Mapping [16.387666608029882]
  We introduce OpenEarthMap-SAR, a benchmark SAR dataset for global high-resolution land cover mapping.<n>OpenEarthMap-SAR consists of 1.5 million segments of 5033 aerial and satellite images with the size of 1024$times$1024 pixels, covering 35 regions from Japan, France, and the USA.<n>We evaluate the performance of state-of-the-art methods for semantic segmentation and present challenging problem settings suitable for further technical development.
  arXiv  Detail & Related papers  (2025-01-18T22:30:27Z)
• GEOBench-VLM: Benchmarking Vision-Language Models for Geospatial Tasks [84.86699025256705]
  We present GEOBench-VLM, a benchmark specifically designed to evaluate Vision-Language Models (VLMs) on geospatial tasks.<n>Our benchmark features over 10,000 manually verified instructions and spanning diverse visual conditions, object types, and scales.<n>We evaluate several state-of-the-art VLMs to assess performance on geospatial-specific challenges.
  arXiv  Detail & Related papers  (2024-11-28T18:59:56Z)
• SpectralEarth: Training Hyperspectral Foundation Models at Scale [47.93167977587301]
  We introduce SpectralEarth, a large-scale multi-temporal dataset designed to pretrain hyperspectral foundation models. We pretrain a series of foundation models on SpectralEarth using state-of-the-art self-supervised learning (SSL) algorithms. We construct four downstream datasets for land-cover and crop-type mapping, providing benchmarks for model evaluation.
  arXiv  Detail & Related papers  (2024-08-15T22:55:59Z)
• Country-Scale Cropland Mapping in Data-Scarce Settings Using Deep Learning: A Case Study of Nigeria [0.6827423171182154]
  We combine a global cropland dataset and a hand-labeled dataset to train machine learning models for generating a new cropland map for Nigeria in 2020 at 10 m resolution. We provide the models with pixel-wise time series input data from remote sensing sources such as Sentinel-1 and 2, ERA5 climate data, and DEM data, in addition to binary labels indicating cropland presence. We find that the existing WorldCover map performs the best with an F1-score of 0.825 and accuracy of 0.870 on the test set, followed by a single-headed LSTM model trained with our hand-labeled training
  arXiv  Detail & Related papers  (2023-12-18T01:23:22Z)
• GeoLLM: Extracting Geospatial Knowledge from Large Language Models [49.20315582673223]
  We present GeoLLM, a novel method that can effectively extract geospatial knowledge from large language models. We demonstrate the utility of our approach across multiple tasks of central interest to the international community, including the measurement of population density and economic livelihoods. Our experiments reveal that LLMs are remarkably sample-efficient, rich in geospatial information, and robust across the globe.
  arXiv  Detail & Related papers  (2023-10-10T00:03:23Z)
• SSL4EO-L: Datasets and Foundation Models for Landsat Imagery [8.34029977985994]
  The Landsat program is the longest-running Earth observation program in history, with 50+ years of data acquisition by 8 satellites. Despite the increasing popularity of deep learning and remote sensing, the majority of researchers still use decision trees and random forests for Landsat image analysis. This paper introduces SSL4EO-L, the first ever dataset designed for Self-Supervised Learning for Earth Observation for the Landsat family of satellites.
  arXiv  Detail & Related papers  (2023-06-15T18:11:20Z)
• SATIN: A Multi-Task Metadataset for Classifying Satellite Imagery using Vision-Language Models [33.814335088752046]
  We introduce SATellite ImageNet (SATIN), a metadataset curated from 27 existing remotely sensed datasets. We comprehensively evaluate the zero-shot transfer classification capabilities of a broad range of vision-language (VL) models on SATIN. We find SATIN to be a challenging benchmark-the strongest method we evaluate achieves a classification accuracy of 52.0%.
  arXiv  Detail & Related papers  (2023-04-23T11:23:05Z)
• 4Seasons: Benchmarking Visual SLAM and Long-Term Localization for Autonomous Driving in Challenging Conditions [54.59279160621111]
  We present a novel visual SLAM and long-term localization benchmark for autonomous driving in challenging conditions based on the large-scale 4Seasons dataset. The proposed benchmark provides drastic appearance variations caused by seasonal changes and diverse weather and illumination conditions. We introduce a new unified benchmark for jointly evaluating visual odometry, global place recognition, and map-based visual localization performance.
  arXiv  Detail & Related papers  (2022-12-31T13:52:36Z)
• A Coarse-to-Fine Approach for Urban Land Use Mapping Based on Multisource Geospatial Data [4.2968261363970095]
  We propose a machine learning-based approach for parcel-level urban land use mapping. We first divide the city into built-up and non-built-up regions based on parcels generated from road networks. We then adopt different classification strategies for parcels in different regions, and finally combine the classified results into an integrated land use map.
  arXiv  Detail & Related papers  (2022-08-18T13:30:56Z)
• Improving Visual Grounding by Encouraging Consistent Gradient-based Explanations [58.442103936918805]
  We show that Attention Mask Consistency produces superior visual grounding results than previous methods. AMC is effective, easy to implement, and is general as it can be adopted by any vision-language model.
  arXiv  Detail & Related papers  (2022-06-30T17:55:12Z)
• Embedding Earth: Self-supervised contrastive pre-training for dense land cover classification [61.44538721707377]
  We present Embedding Earth a self-supervised contrastive pre-training method for leveraging the large availability of satellite imagery. We observe significant improvements up to 25% absolute mIoU when pre-trained with our proposed method. We find that learnt features can generalize between disparate regions opening up the possibility of using the proposed pre-training scheme.
  arXiv  Detail & Related papers  (2022-03-11T16:14:14Z)
• Jalisco's multiclass land cover analysis and classification using a novel lightweight convnet with real-world multispectral and relief data [51.715517570634994]
  We present our novel lightweight (only 89k parameters) Convolution Neural Network (ConvNet) to make LC classification and analysis. In this work, we combine three real-world open data sources to obtain 13 channels. Our embedded analysis anticipates the limited performance in some classes and gives us the opportunity to group the most similar.
  arXiv  Detail & Related papers  (2022-01-26T14:58:51Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.