Moonstone: A Multimodal Foundation Model and Benchmark for Lunar Remote Sensing

• URL: http://arxiv.org/abs/2607.03644v1
• Date: Fri, 03 Jul 2026 23:54:52 GMT
• Title: Moonstone: A Multimodal Foundation Model and Benchmark for Lunar Remote Sensing
• Abstract summary: We introduce Moonstone, the first multi-modal foundation model benchmark for lunar remote sensing.<n>Our contributions are: (1) a 28-channel, 128 pixels-per-degree (237 m) global lunar pretraining dataset from seven instrument families across five missions, (2) MG-MAE, a modality-grouped masked autoencoder with per-group convolutional tokenizers, and (3) a benchmark of six downstream tasks covering classification, regression, and segmentation.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Decades of orbital missions have produced multi-modal remote sensing data for the Moon, spanning optical imagery, spectroscopy, thermal emission, radar, gravity, and elemental composition. Yet these datasets remain fragmented across archives, and no benchmark exists for evaluating machine learning on lunar data. We introduce Moonstone, the first multi-modal foundation model benchmark for lunar remote sensing. Our contributions are: (1) a 28-channel, 128 pixels-per-degree (~237 m) global lunar pretraining dataset from seven instrument families across five missions, (2) MG-MAE, a modality-grouped masked autoencoder with per-group convolutional tokenizers, a shared Vision Transformer encoder, attention masking for missing modalities, coverage-adaptive masking for heterogeneous spatial coverage, and spectral continuity regularization for physically plausible reconstructions, and (3) a benchmark of six downstream tasks covering classification, regression, and segmentation. MG-MAE pretrained features outperform scratch baselines on all tasks and surpass both ImageNet-pretrained and vanilla MAE baselines by large margins. Data and code are available at https://huggingface.co/datasets/ayushprd/Moonstone and https://github.com/ayushprd/Moonstone .

Related papers

• SomBench: Benchmark Dataset for Advancing Machine Learning in Lunar Science [4.888910677829709]
  We introduce SomBench, a unified, spatially-aligned, ML-ready lunar dataset aggregating 30+ co-registered layers from ten instruments across four missions.<n>An application benchmark suite spans impact processes, volcanic history, and polar volatiles.<n> Baseline experiments with ResNet-50 and SwinV2-B models confirm that each benchmark task is learnable from the released inputs.
  arXiv  Detail & Related papers  (2026-09-08T02:58:32Z)
• LunarFM: A Shared Multimodal Representation of the Moon's Surface [3.029669084751708]
  We introduce LunarFM, a model that learns a general representation of the lunar surface from diverse orbital measurements.<n>LunarFM assimilates observations from six instruments across three lunar missions, mapping 18 input channels to a shared embedding space.<n>We demonstrate that this embedding space supports a diverse range of downstream applications, including similarity search, few-shot resource mapping, mineral abundance regression, and geological unit classification.
  arXiv  Detail & Related papers  (2026-07-24T15:25:20Z)
• Towards UAV Detection in the Real World: A New Multispectral Dataset UAVNet-MS and a New Method [80.50361258830928]
  UAVNet-MS is the first multispectral dataset for fine-grained small-UAV detection.<n>We propose MFDNet, a dual-stream baseline addressing array-induced parallax and spatial-spectral fusion.<n>This work provides foundational dataset, strong baseline, and benchmark for multispectral UAV monitoring research.
  arXiv  Detail & Related papers  (2026-05-20T09:49:04Z)
• VFM-Recon: Unlocking Cross-Domain Scene-Level Neural Reconstruction with Scale-Aligned Foundation Priors [49.39553550491549]
  We present VFMRecon, the first attempt to bridge transferable VFM priors with scaleconsistent requirements in scene-level neural reconstruction.<n> Specifically, we first introduce a lightweight scale alignment stage that restores multiview scale coherence.<n>We then integrate pretrained VFM features into the neural volumetric reconstruction pipeline via lightweight task-specific adapters.
  arXiv  Detail & Related papers  (2026-03-13T05:00:44Z)
• SpaceSense-Bench: A Large-Scale Multi-Modal Benchmark for Spacecraft Perception and Pose Estimation [1.0262304700896199]
  We present textbfSpaceSense-Bench, a large-scale multi-modal benchmark for spacecraft perception.<n>Each frame provides time-synchronized 1024$times$1024 RGB images, millimeter-precision depth maps, and 256-beam LiDAR point clouds.<n>We benchmark five representative tasks (object detection, 2D semantic segmentation, RGB--LiDAR fusion-based 3D point cloud segmentation, monocular depth estimation, and orientation estimation) and identify two key findings.
  arXiv  Detail & Related papers  (2026-03-10T07:52:28Z)
• LunarLoc: Segment-Based Global Localization on the Moon [52.418394025548295]
  Global localization is necessary for autonomous operations on the lunar surface where traditional Earth-based navigation infrastructure, such as GPS, is unavailable.<n>We propose LunarLoc, an approach to global localization that leverages instance segmentation for zero-shot extraction of boulder landmarks from onboard stereo imagery.<n>LunarLoc achieves sub-cm level accuracy in multi-session global localization experiments, significantly outperforming the state of the art in lunar global localization.
  arXiv  Detail & Related papers  (2025-06-20T12:06:47Z)
• LuSeg: Efficient Negative and Positive Obstacles Segmentation via Contrast-Driven Multi-Modal Feature Fusion on the Lunar [8.215362367428565]
  We have developed a lunar surface simulation system called the Lunar Exploration Simulator System (LESS)<n>We also propose a novel two-stage segmentation network called LuSeg.<n>LuSeg enforces semantic consistency between the RGB encoder from Stage I and the depth from Stage II.
  arXiv  Detail & Related papers  (2025-03-14T13:51:52Z)
• EarthView: A Large Scale Remote Sensing Dataset for Self-Supervision [72.84868704100595]
  This paper presents a dataset specifically designed for self-supervision on remote sensing data, intended to enhance deep learning applications on Earth monitoring tasks.<n>The dataset spans 15 tera pixels of global remote-sensing data, combining imagery from a diverse range of sources, including NEON, Sentinel, and a novel release of 1m spatial resolution data from Satellogic.<n>Accompanying the dataset is EarthMAE, a tailored Masked Autoencoder developed to tackle the distinct challenges of remote sensing data.
  arXiv  Detail & Related papers  (2025-01-14T13:42:22Z)
• PolSAM: Polarimetric Scattering Mechanism Informed Segment Anything Model [83.35198885088093]
  PolSAR data presents unique challenges due to its rich and complex characteristics.<n>Existing data representations, such as complex-valued data, polarimetric features, and amplitude images, are widely used.<n>Most feature extraction networks for PolSAR are small, limiting their ability to capture features effectively.<n>We propose the Polarimetric Scattering Mechanism-Informed SAM (PolSAM), an enhanced Segment Anything Model (SAM) that integrates domain-specific scattering characteristics and a novel prompt generation strategy.
  arXiv  Detail & Related papers  (2024-12-17T09:59:53Z)
• LuSNAR:A Lunar Segmentation, Navigation and Reconstruction Dataset based on Muti-sensor for Autonomous Exploration [2.3011380360879237]
  Environmental perception and navigation algorithms are the foundation for lunar rovers. Most of the existing lunar datasets are targeted at a single task. We propose a multi-task, multi-scene, and multi-label lunar benchmark dataset LuSNAR.
  arXiv  Detail & Related papers  (2024-07-09T02:47:58Z)
• MMScan: A Multi-Modal 3D Scene Dataset with Hierarchical Grounded Language Annotations [55.022519020409405]
  This paper builds the first largest ever multi-modal 3D scene dataset and benchmark with hierarchical grounded language annotations, MMScan.<n>The resulting multi-modal 3D dataset encompasses 1.4M meta-annotated captions on 109k objects and 7.7k regions as well as over 3.04M diverse samples for 3D visual grounding and question-answering benchmarks.
  arXiv  Detail & Related papers  (2024-06-13T17:59:30Z)

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.