datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
TreeSatAI-Time-Series
TreeSatAI-Time-Series
This dataset was introduced in the ECCV24 paper OmniSat.
Ahlswede et al. (https://essd.copernicus.org/articles/15/681/2023/) introduced the TreeSatAI Benchmark Archive, a new dataset for tree species classification in Central Europe based on multi-sensor data from aerial,
Sentinel-1 and Sentinel-2. The dataset contains labels of 20 European tree species (i.e., 15 tree genera) derived from forest administration data of the federal state of Lower Saxony… See the full description on the dataset page: https://huggingface.co/datasets/IGNF/TreeSatAI-Time-Series.Raw-LCD-Utility-Meter-Dial-and-Serial-Number-Images
Dual-Target Digital LCD Utility Meter Images — Sample Preview
This preview showcases 50 sample images from our full-scale dataset Images_electro_scor_ser_v1_2026 — a collection of 2,290 high-resolution, real-world images of utility meters featuring both electronic (LCD/LED) displays and their manufacturer serial barcodes.
Dataset Details
Built for Computer Vision teams training automated meter-reading systems, this dataset goes beyond generic data by capturing the… See the full description on the dataset page: https://huggingface.co/datasets/utilitimetersai/Raw-LCD-Utility-Meter-Dial-and-Serial-Number-Images.Raw-Mechanical-Utility-Meter-Dial-and-Serial-Number-Images
Raw Mechanical Utility Meter Dial & Serial Number Images (Preview Set – 50 Images)
Dual-Target Real-World Meter Photos for Computer Vision & OCR Training
Billions of legacy mechanical utility meters are still running worldwide, and teaching a model to read spinning analog dials alongside engraved serial numbers requires serious visual diversity. This preview release gives you 50 free sample images pulled from our full 9,299-image mechanical meter collection, so you can evaluate… See the full description on the dataset page: https://huggingface.co/datasets/utilitimetersai/Raw-Mechanical-Utility-Meter-Dial-and-Serial-Number-Images.Raw-Mix-Mechanical-LCD-Meter-Dial-and-Serial-Number-Images
Ultimate Mix: Mechanical + LCD/LED Utility Meter Images (Preview Set – 50 Images)
50 real-world images free — full 11,589-image dataset available via UtilityMeters.ai
Train a single, universal detection model that works across legacy mechanical dials and modern LCD/LED smart meters — no separate pipelines needed.
This free sample gives you 50 unannotated, high-fidelity raw images pulled directly from our full 11,589-image "Ultimate Mix" collection, covering electricity, water… See the full description on the dataset page: https://huggingface.co/datasets/utilitimetersai/Raw-Mix-Mechanical-LCD-Meter-Dial-and-Serial-Number-Images.passport-faces
Passport Faces
Face crop images organized into men/ and women/ folders, with a _metadata.csv in each.
motor-nameplate
Motor Nameplate
A small image dataset of electric motor nameplates collected from public Google Images results, intended for tasks such as:
Training OCR / document-understanding models to extract nameplate text fields (manufacturer, HP, RPM, voltage, frame, etc.)
Image classification by manufacturer (ABB, Siemens, Baldor-Reliance, WEG, Hyundai, etc.)
Object detection (locating the plate on the motor body)
Few-shot learning / evaluation baselines for industrial vision tasks
The… See the full description on the dataset page: https://huggingface.co/datasets/sergiudanstan/motor-nameplate.salmonella-serovar-hyperspectral
Salmonella Serovar Hyperspectral Microscopy (Foods 2025)
Salmonella Serovar Hyperspectral Microscopy is an image dataset for foodborne bacterial classification using hyperspectral imaging. It was created to support research in rapid pathogen identification, enabling models to classify Salmonella serovars directly from microscopy images without the need for selective enrichment.
Companion spectral dataset: The single-cell spectral features (tabular, 25,972 rows) extracted from these… See the full description on the dataset page: https://huggingface.co/datasets/food-ai-nexus/salmonella-serovar-hyperspectral.KIDOKIDO: Emotion Analysis of Children’s Drawings
About
KIDO is a novel dataset for analyzing how children express emotions through hand-drawn images and self-reported reflections.
It contains 5,430 children’s drawings, each paired with text indicating whether the drawing represents happiness or sadness.
Unlike prior datasets that focus only on digital drawings from tablets, KIDO captures natural scanned hand-drawn expressions.
Details can be found in our paper:… See the full description on the dataset page: https://huggingface.co/datasets/serdarciftci/KIDO.signatures
Signatures
1176 cropped handwritten signature images (PNG, RGBA with transparent background), tightly cropped to the ink with a small uniform margin.
Files are named sig_0001.png ... sig_1176.png.
food-waste-dataset
Food Waste Dataset
This dataset contains detailed information about food waste, including images and nutritional information.
Dataset Description
The dataset combines image data with detailed nutritional information for each meal and its ingredients.
Features
For each entry:
bonid: Unique identifier for each meal
image: Image of the meal
Lists per meal (multiple ingredients):
Artikelnummer: Article numbers
Artikel: Ingredient names
Stückartikel: Piece article… See the full description on the dataset page: https://huggingface.co/datasets/AI-ServicesBB/food-waste-dataset.SERGE
SERGE (SEmantic Richness & Geometric Evaluation)
SERGE is the evaluation dataset released with the paper Semantic Richness or Geometric Reasoning? The Fragility of VLM's Visual Invariance. It contains the canonical (untransformed) images used to test whether VLMs judge object identity under rotation, scale, and identity transformations through geometric reasoning or through familiarity with semantically rich content.
Paper: arXiv:2604.01848
Project page:… See the full description on the dataset page: https://huggingface.co/datasets/XThomasBU/SERGE.Synthetic-X-Ray-Dataset
Self-Supervised Learning Powered by Synthetic Data From Diffusion Models: Unconditional Synthetic Images for Chest Xray
This repository contains a fully synthetic dataset used for the study, Self-Supervised Learning Powered by Synthetic Data From Diffusion Models: Application to X-Ray Images.
Abstract:
Synthetic data offers a compelling solution to the challenges associated with acquiring high-quality medical data, which is often constrained by privacy concerns and limited… See the full description on the dataset page: https://huggingface.co/datasets/serag-ai/Synthetic-X-Ray-Dataset.ser_balanced
Snapshot Safari SER (Serengeti) — Balanced Classroom Subset v1.0
Summary
A curated, balanced subset of the Snapshot Safari 2024 Expansion SER (Serengeti
National Park) camera trap dataset, prepared for use in the CAS Deep Learning —
Computer Vision course exercises.
Archive
Images
Description
ser_balanced.tar.gz
1,850
Balanced, ≤200/class, full frames
Source
Dataset: Snapshot Safari 2024 Expansion — SER (Serengeti) subset
URL:… See the full description on the dataset page: https://huggingface.co/datasets/marco-willi/ser_balanced.SerreOlivaPoggio-PNAS07
Animal / Non-Animal Stimulus Set — Serre, Oliva & Poggio (PNAS 2007)
This dataset is the natural-image stimulus set used in:
Serre, T., Oliva, A., & Poggio, T. (2007). A feedforward architecture accounts for rapid categorization. Proceedings of the National Academy of Sciences, 104(15), 6424–6429. https://doi.org/10.1073/pnas.0700622104
It was the basis for the rapid animal/non-animal categorization psychophysics task in that paper, and has since been re-used in many… See the full description on the dataset page: https://huggingface.co/datasets/Serrelab/SerreOlivaPoggio-PNAS07.Anonymized-Maid-Service-V1-Preview
🧹 BWS Maid_Pro: Compliance-Native Multimodal Tokens (V1)
🛡️ Engineering Evaluation Sandbox (Active 7-Day Access)
Technical Ingestion Portal: s3://createphotos (Whitelisted buckets only)
Secure Evaluation Link: Download 11_Maid_Service_30_Enterprise_POC.zip
Direct Manifest Auditor: BWS Forensic Manifest Repository
Procurement: All assets are 2026 US CLEAR Act compliant. Access is granted to whitelisted engineering nodes only. Forward your AWS Account ID to… See the full description on the dataset page: https://huggingface.co/datasets/BWS-Data-Solutions/Anonymized-Maid-Service-V1-Preview.11-Maid-Service-30-Enterprise-CLEAR-Compliant-POC
🧹 BWS Maid Service: Compliance-Native Multimodal Tokens (POC)
🛡️ Engineering Evaluation Sandbox (Active 7-Day Access)
Technical Ingestion Portal: s3://createphotos (Whitelisted buckets only)
Secure Evaluation Link: Download 11_Maid_Service_30_Enterprise_POC.zip
Direct Manifest Auditor: BWS Forensic Manifest Repository
Procurement: All assets are 2026 US CLEAR Act compliant. Access is granted to whitelisted engineering nodes only. Forward your AWS Account ID to… See the full description on the dataset page: https://huggingface.co/datasets/BWS-Data-Solutions/11-Maid-Service-30-Enterprise-CLEAR-Compliant-POC.hazard_info
Global Multi-Hazard Risk Assessment Dataset
Dataset Description
This dataset contains comprehensive global multi-hazard risk assessment data covering 9 different natural hazard types. The data is sourced from the UNEP Grid and represents state-of-the-art hazard modeling for disaster risk assessment and climate change adaptation planning.
Dataset Summary
Total Size: ~50GB of geospatial data
Coverage: Global
Resolution: Variable (1km to 11km depending on hazard… See the full description on the dataset page: https://huggingface.co/datasets/serenia-science/hazard_info.rvl_cdipThe RVL-CDIP (Ryerson Vision Lab Complex Document Information Processing) dataset consists of 400,000 grayscale images in 16 classes, with 25,000 images per class. There are 320,000 training images, 40,000 validation images, and 40,000 test images.