datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
sentineledgedataset
SentinelEdge Traffic Dataset
A large-scale urban traffic dataset combining Bengaluru CCTV footage with the UA-DETRAC benchmark, assembled for the SentinelEdge traffic management and vehicle detection system. Contains ~196,000 annotated images across four sub-datasets covering Indian urban roads and international highway scenarios.
Sub-datasets
1. BMD-45-Train
35,792 images from 45 Bengaluru CCTV locations
373,132 bounding box annotations in COCO… See the full description on the dataset page: https://huggingface.co/datasets/kalyan1729/sentineledgedataset.agent-trajectory-sentinel
AgentTrajectorySentinel — 3581 agent episodes across 33 corpora
Committed agent trajectories with step-level telemetry, used to fit and
evaluate one-class monitors for real-time failure detection.
Paper: https://arxiv.org/abs/2608.02464
Code and the full evaluation harness:
https://github.com/sunnydubey1111/agent-trajectory-sentinel
A recorded walkthrough of the method, ending with the live demo
detecting and repairing a real failure:
https://youtu.be/a05n_000klE?t=0… See the full description on the dataset page: https://huggingface.co/datasets/sunnydubey1111/agent-trajectory-sentinel.Sentinel_Agri_China
Sentinel Patch Imagery Dataset for China's Cropland
1. Dataset Description
Dataset Name: Sentinel Patch Imagery Dataset for China's Cropland
Version: V1.0
Authors: Mo Wang, Jing Wang, Yunpeng Cui, Juan Liu, Li Chen, Ting Wang, Jinming Wu, Huan Li
License: apache-2.0
Examples: chunk_41 (Example_S1_chunk_41.zip.001 and Example_S2_chunk_41.zip.001) serves as example datasets as this chunk is small.
1.1. Overview
This project is inspired by… See the full description on the dataset page: https://huggingface.co/datasets/zerria/Sentinel_Agri_China.sentinel-beetles
Dataset Card for Beetles as Sentinel Taxa: Predicting drought conditions from NEON specimen imagery
This dataset contains images of pinned carabid beetle specimens collected by the National Ecological Observatory Network (NEON) from ecological sites across the U.S., along with associated metadata and drought severity indices (Standardized Precipitation Evapotranspiration Index (SPEI)). It was developed to for the second HDR ML Challenge, and is intended to support the development of… See the full description on the dataset page: https://huggingface.co/datasets/imageomics/sentinel-beetles.amfitrite-inland-waters-hab-sentinel2
Dataset Card for Amfitrite-Inland-Waters-HAB-Sentinel2
This dataset contains multispectral Sentinel-2 satellite imagery tiles focused on inland water bodies, classified by the severity of Harmful Algal Blooms (HABs).
It is designed to train Deep Learning models (like CNNs) for environmental monitoring.
Dataset Details
Dataset Description
Amfitrite-Inland-Waters-HAB-Sentinel2 is a specialized dataset designed for the detection and classification of… See the full description on the dataset page: https://huggingface.co/datasets/kostaspic/amfitrite-inland-waters-hab-sentinel2.alignment-seeds
Sentinel Alignment Seeds v2
Text is risk. Action is danger. Sentinel watches both.
Validated alignment seeds for LLMs and autonomous agents. Prompt-based safety that works without model modification.
What's New in v2
The THSP Protocol — A four-gate evaluation system:
Truth — No deception
Harm — No damage
Scope — Within boundaries
Purpose — Must serve legitimate benefit ← NEW
The PURPOSE gate catches actions that pass harm-avoidance but serve no one (e.g., "drop the… See the full description on the dataset page: https://huggingface.co/datasets/sentinelseed/alignment-seeds.sentinel-2-reanalysis-NDVI-EVI-TCG
Description
This dataset contains Sentinel-2 raw bands and reanalysis (NDVI, EVI, TCG) with valid pixel fractions per field.
Each parquet file contains multiple sunflower crop fields (identified as field_id).
Each field has 36 time stamps per year which is 10-day interval aggregation of the values of scaled and filtered bands and reanalysis.
Shape files of field polygons were loaded from Google Earth Engine (GEE) project assets.
Content
Country: Hungary
Fields: Sunflower… See the full description on the dataset page: https://huggingface.co/datasets/Delineeeee/sentinel-2-reanalysis-NDVI-EVI-TCG.sentinelng-data-crop-corn
SentinelNG Corn Crop Dataset
This repository is a corn-focused image dataset for SentinelNG crop-health or disease-classification experiments. The published tree is organized as image-folder data with corn-related class directories.
Recommended loading and evaluation
Use an image-folder loader after checking the class names and image quality. Keep images from the same field, plant, or capture session in a single split where possible. Report per-class precision… See the full description on the dataset page: https://huggingface.co/datasets/MR-CODESPIKE/sentinelng-data-crop-corn.SENTINEL
Dataset Card for ICCV 2025 | SENTINEL:Mitigating Object Hallucinations via Sentence-Level Early Intervention
Paper Abstract
Multimodal large language models (MLLMs) have revolutionized cross-modal understanding but continue to struggle with hallucinations - fabricated content contradicting visual inputs. Existing hallucination mitigation methods either incur prohibitive computational costs or introduce distribution mismatches between training data… See the full description on the dataset page: https://huggingface.co/datasets/psp-dada/SENTINEL.sentinelng-data-crop
SentinelNG Crop Image Dataset
This repository contains image-folder resources for crop or plant-disease classification in the SentinelNG project. Use the directory structure and class folders as the source of truth for the currently published categories.
Reproducible use
Inspect image dimensions, formats, duplicate content, and class distribution before training. Split by plant, field, or source collection where that information exists, rather than relying only on… See the full description on the dataset page: https://huggingface.co/datasets/MR-CODESPIKE/sentinelng-data-crop.Uniform-Sentinel-1-2-Dataset
Dataset name
UNIFORM SEN1-2
Year of publication
2025
Author
Shamba Chowdhury
Ankana Ghosh
Shreyashi Ghosh
License
CC-BY-SA-4.0
The dataset contains Copernicus data (2024). Terms and conditions apply: https://scihub.copernicus.eu/twiki/pub/SciHubWebPortal/TermsConditions/TC_Sentinel_Data_31072014.pdf
Associated publication
TBA
Links
Dataset: https://www.kaggle.com/datasets/shambac/uniform-sentinel-1-2-dataset
Paper: TBA… See the full description on the dataset page: https://huggingface.co/datasets/ShambaC/Uniform-Sentinel-1-2-Dataset.Global-Spatiotemporal-Sentinel-2-Dataset-with-Pre-computed-GEE-Embeddings-2017-2024
Citation
@ARTICLE{11589351,
author={Kömürcü, Kürşat and Petkevicius, Linas},
journal={IEEE Geoscience and Remote Sensing Letters},
title={Sat-JEPA-Diff: Caption-Guided Zero-RGB Satellite Image Forecasting via Self-Supervised Diffusion},
year={2026},
volume={},
number={},
pages={1-1},
doi={10.1109/LGRS.2026.3708464}
}
@inproceedings{
komurcu2026satjepadiff,
title={Sat-{JEPA}-Diff: Bridging Self-Supervised Learning and Generative Diffusion for Remote… See the full description on the dataset page: https://huggingface.co/datasets/kursatkomurcu/Global-Spatiotemporal-Sentinel-2-Dataset-with-Pre-computed-GEE-Embeddings-2017-2024.sentinel-2-reanalysis-NDVI-EVI-TCG
Description
This dataset contains Sentinel-2 raw bands and reanalysis (NDVI, EVI, TCG) with valid pixel fractions per field.
Each parquet file contains multiple sunflower crop fields (identified as field_id).
Each field has 36 time stamps per year which is 10-day interval aggregation of the values of scaled and filtered bands and reanalysis.
Shape files of field polygons were loaded from Google Earth Engine (GEE) project assets.
Content
Country: Hungary
Fields: Sunflower… See the full description on the dataset page: https://huggingface.co/datasets/jaelin215/sentinel-2-reanalysis-NDVI-EVI-TCG.wildlife-sentinel
Wildlife Sentinel
Wildlife Sentinel is the public Hugging Face dataset package for the Savanna Sentinel project: a multimodal, tool-grounded biodiversity monitoring dataset built from Snapshot Serengeti camera-trap events and public environmental layers.
This Hub package intentionally contains the train, validation, and evaluation components only, plus supporting artifacts needed to understand and reproduce those splits. Demo replay inputs, hidden demo labels, raw source caches… See the full description on the dataset page: https://huggingface.co/datasets/Alfaxad/wildlife-sentinel.sentinel_strawberries_v2This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "sentinel_v2",
"total_episodes": 1,
"total_frames": 465,
"total_tasks": 1,
"total_videos": 0,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 20,
"splits": {
"train": "0:1"
},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/grboguz/sentinel_strawberries_v2.amfitrite-open-waters-hab-sentinel2
Dataset Card for Amfitrite-Open-Waters-HAB-Sentinel2
This dataset contains multispectral Sentinel-2 satellite imagery tiles focused on open water and coastal marine environments, classified by the potential presence of Harmful Algal Blooms (HABs).
It is designed to train Machine or Deep Learning models (like CNNs) for large-scale environmental monitoring and ocean anomaly detection.
Dataset Details
Dataset Description
Amfitrite-Open-Waters-HAB-Sentinel2 is a… See the full description on the dataset page: https://huggingface.co/datasets/kostaspic/amfitrite-open-waters-hab-sentinel2.gargano-sentinel
Gargano Sentinel-2 (clear-sky)
Cloud-filtered, fully-tiled Sentinel-2 acquisitions over the Gargano peninsula
(Apulia, Italy). Each acquisition is split into a regular grid of 40 tiles
(5 rows × 8 columns) covering roughly 10 × 10 km each at
1000 × 1000 px.
Only complete acquisitions (all 40 tiles available from a single overpass)
whose mean tile cloud fraction is ≤ 0.5 are included.
Splits
The dataset is organised in three splits, one per band combination. Every… See the full description on the dataset page: https://huggingface.co/datasets/MikCil/gargano-sentinel.sentinel-2-rgb-captionedGitHub: https://github.com/sshh12/terrain-diffusion
Source
COPERNICUS_S2_SR
Contains modified Copernicus Service information 2023
Captions
Generated using openai/clip-vit-large-patch14 on a fixed set of captions.
Data Processing
Pick region, filter for low clouds, get median over fixed timespan
Normalize RGB by min/max channel values
Generate captions for each tile using CLIP
Filter out tiles with black parts or overlap lines (automatically)
Manually pass to… See the full description on the dataset page: https://huggingface.co/datasets/sshh12/sentinel-2-rgb-captioned.le2i-sentinel-frames
Le2i Sentinel Test Frames
Extracted frames from the Le2i Fall Detection Dataset for benchmarking the Sentinel video intelligence pipeline.
Contents
130 annotated videos (99 falls, 31 normals)
3 frames per video (390 total)
4 environments: Coffee_room_01, Coffee_room_02, Home_01, Home_02
Ground truth in metadata/ground_truth.json
Source
Le2i Fall Detection Dataset (University of Burgundy)
Resolution: 320x240 @ 25fps
Citation: Charfi et al., "Optimised… See the full description on the dataset page: https://huggingface.co/datasets/seanphan/le2i-sentinel-frames.sentinel_dataset_v15_deltaThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "sentinel_v2",
"total_episodes": 1,
"total_frames": 484,
"total_tasks": 1,
"total_videos": 2,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 15,
"splits": {
"train": "0:1"
},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/grboguz/sentinel_dataset_v15_delta.sentinel_strawberries_v6This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "sentinel_v2",
"total_episodes": 1,
"total_frames": 119,
"total_tasks": 1,
"total_videos": 2,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 20,
"splits": {
"train": "0:1"
},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/grboguz/sentinel_strawberries_v6.SentinelKilnDB
SentinelKilnDB - A Large-Scale Dataset and Benchmark for OBB Brick Kiln Detection in South Asia Using Satellite Imagery
NeurIPS 2025 Datasets & Benchmarks Track
Abstract
Air pollution was responsible for 2.6 million deaths across South Asia in 2021 alone, with brick manufacturing contributing significantly to this burden. In particular, the Indo-Gangetic Plain; a densely populated and highly polluted region spanning northern India, Pakistan, Bangladesh, and… See the full description on the dataset page: https://huggingface.co/datasets/SustainabilityLabIITGN/SentinelKilnDB.sentinelng-data-health
SentinelNG Health Image Dataset
This dataset contains image-folder resources for health-related classification or screening experiments in SentinelNG. The repository is organized as image data rather than application code.
Use and handling
Treat the directory labels as class identifiers only after inspecting the complete tree and associated provenance. Build explicit train/validation/test splits that prevent near-duplicate or patient-linked images from crossing… See the full description on the dataset page: https://huggingface.co/datasets/MR-CODESPIKE/sentinelng-data-health.golden-batch-sentinel-data
Golden Batch Sentinel Data
Benchmark datasets for process monitoring and fault detection in batch manufacturing.
Datasets
IndPenSim (Industrial Penicillin Simulation)
A 100,000L fermentation simulation with 100 batches and rich multivariate signals.
Source: Mendeley Data
Paper: Modern day monitoring and control challenges...
Batches: 100 (90 normal, 10 faulty)
Variables: 37 process variables (Raman spectra excluded for efficiency)
Time resolution: 0.2 hours… See the full description on the dataset page: https://huggingface.co/datasets/foundation-models/golden-batch-sentinel-data.sentinel-jailbreak-detectionSentinelAI-Featuressentinel_trajectory_ist_montsentinel2-lejepa-global-diverse-256
Sentinel-2 LeJEPA Preset-Biased (Small)
A small, preset-biased Sentinel-2 L2A chip dataset curated for self-supervised
pretraining of a LeJEPA ResNet-18 encoder.
Built as a reproducibility artifact for the
Sentinel Change Explorer
proof-of-concept foundation-model change-detection feature.
This is a proof of concept, not a general-purpose EO pretraining corpus. It
is intentionally tiny (~thousands of chips) and biased toward the five demo
AOIs the Sentinel Change Explorer app… See the full description on the dataset page: https://huggingface.co/datasets/falafel-hockey/sentinel2-lejepa-global-diverse-256.sentinel-2-forest-browning-monitoring-app-storagesentinel-security-bot