datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
cifar10
Dataset Card for CIFAR-10
Dataset Summary
The CIFAR-10 dataset consists of 60000 32x32 colour images in 10 classes, with 6000 images per class. There are 50000 training images and 10000 test images.
The dataset is divided into five training batches and one test batch, each with 10000 images. The test batch contains exactly 1000 randomly-selected images from each class. The training batches contain the remaining images in random order, but some training batches may contain… See the full description on the dataset page: https://huggingface.co/datasets/uoft-cs/cifar10.typed_digital_signatures
Typed Digital Signatures Dataset
This comprehensive dataset contains synthetic digital signatures rendered across 30 different Google Fonts, specifically selected for their handwriting and signature-style characteristics. Each font contributes unique stylistic elements, making this dataset ideal for robust signature analysis and font recognition tasks.
Dataset Overview
Total Fonts: 30 different Google Fonts
Images per Font: 3,000 signatures
Total Dataset Size:… See the full description on the dataset page: https://huggingface.co/datasets/Benjy/typed_digital_signatures.imagenet-1k
Dataset Card for ImageNet
Dataset Summary
ILSVRC 2012, commonly known as 'ImageNet' is an image dataset organized according to the WordNet hierarchy. Each meaningful concept in WordNet, possibly described by multiple words or word phrases, is called a "synonym set" or "synset". There are more than 100,000 synsets in WordNet, majority of them are nouns (80,000+). ImageNet aims to provide on average 1000 images to illustrate each synset. Images of each concept are… See the full description on the dataset page: https://huggingface.co/datasets/ILSVRC/imagenet-1k.mnist
Dataset Card for MNIST
Dataset Summary
The MNIST dataset consists of 70,000 28x28 black-and-white images of handwritten digits extracted from two NIST databases. There are 60,000 images in the training dataset and 10,000 images in the validation dataset, one class per digit so a total of 10 classes, with 7,000 images (6,000 train images and 1,000 test images) per class.
Half of the image were drawn by Census Bureau employees and the other half by high school students… See the full description on the dataset page: https://huggingface.co/datasets/ylecun/mnist.cifar100
Dataset Card for CIFAR-100
Dataset Summary
The CIFAR-100 dataset consists of 60000 32x32 colour images in 100 classes, with 600 images
per class. There are 500 training images and 100 testing images per class. There are 50000 training images and 10000 test images. The 100 classes are grouped into 20 superclasses.
There are two labels per image - fine label (actual class) and coarse label (superclass).
Supported Tasks and Leaderboards
image-classification: The… See the full description on the dataset page: https://huggingface.co/datasets/uoft-cs/cifar100.food101
Dataset Card for Food-101
Dataset Summary
This dataset consists of 101 food categories, with 101'000 images. For each class, 250 manually reviewed test images are provided as well as 750 training images. On purpose, the training images were not cleaned, and thus still contain some amount of noise. This comes mostly in the form of intense colors and sometimes wrong labels. All images were rescaled to have a maximum side length of 512 pixels.
Supported Tasks and… See the full description on the dataset page: https://huggingface.co/datasets/ethz/food101.cholect50
CholecT50 Dataset (FiftyOne Format)
This is a FiftyOne dataset version of the CholecT50 dataset.
CholecT50 is a dataset of laparoscopic cholecystectomy surgeries, annotated with surgical action triplets. It is the first public dataset to provide action triplet annotations for surgical videos, enabling research in fine-grained surgical activity recognition.
Dataset Summary
CholecT50 consists of 50 videos of laparoscopic cholecystectomy surgeries. Each frame is annotated… See the full description on the dataset page: https://huggingface.co/datasets/Voxel51/cholect50.PhysicalAI-SimReady-Warehouse-01
NVIDIA Physical AI SimReady Warehouse OpenUSD Dataset
Dataset Version: 1.1.0
Date: May 18, 2025
Author: NVIDIA, Corporation
License: CC-BY-4.0 (Creative Commons Attribution 4.0 International)
Contents
This dataset includes the following:
This README file
A CSV catalog that enumerates all of the OpenUSD assets that are part of this dataset including a sub-folder of images that showcase each 3D asset (physical_ai_simready_warehouse_01.csv). The CSV file is organized in… See the full description on the dataset page: https://huggingface.co/datasets/nvidia/PhysicalAI-SimReady-Warehouse-01.GAMMA
GAMMA — Glaucoma grading from Multi-Modality imAges (Challenge dataset)
Image: Dataset Samples.
Short description
GAMMA is the first public multi-modality glaucoma grading dataset that pairs 2D color fundus photographs with 3D OCT volumes for each sample. It was released as part of the GAMMA challenge (OMIA8 / MICCAI 2021) to encourage algorithms that combine fundus and OCT information for automatic… See the full description on the dataset page: https://huggingface.co/datasets/ctmedtech/GAMMA.TreasureEnglish | 中文
News
Our work has been accepted by ICML 2026 as a poster. 🎉🎉
The paper has released on ArXiv.
The code has released on Github.
Treasure Dataset
Treasure is the dataset companion to Fleet (Fleet: Few Shots Lead Effective AI-generated Images Detection), designed for AI-generated image detection in an open-world setting where generators evolve and distribution shift is central. Unlike saturated results on closed datasets, Treasure/Fleet emphasizes… See the full description on the dataset page: https://huggingface.co/datasets/ThreeLiu/Treasure.PASTIS-HD
🌱 PASTIS-HD 🌿 Panoptic Agricultural Satellite TIme Series : optical time series, radar time series and very high resolution image
PASTIS is a benchmark dataset for panoptic and semantic segmentation of agricultural parcels from satellite time series.
It contains 2,433 patches within the French metropolitan territory with panoptic annotations (instance index + semantic label for each pixel).
Each patch is a Sentinel-2 multispectral image time series of variable lentgh.
This… See the full description on the dataset page: https://huggingface.co/datasets/IGNF/PASTIS-HD.tiny-imagenet
Dataset Card for tiny-imagenet
Dataset Summary
Tiny ImageNet contains 100000 images of 200 classes (500 for each class) downsized to 64×64 colored images. Each class has 500 training images, 50 validation images, and 50 test images.
Languages
The class labels in the dataset are in English.
Dataset Structure
Data Instances
{
'image': <PIL.JpegImagePlugin.JpegImageFile image mode=RGB size=64x64 at 0x1A800E8E190,
'label': 15
}… See the full description on the dataset page: https://huggingface.co/datasets/zh-plus/tiny-imagenet.OpenFake
Dataset Card for OpenFake
Known issues
Prompt–image misalignment in the synthetic split (reported November 2025, fix pending)
For five of the eighty generators, the prompt field attached to synthetic
images does not correspond to the prompt actually used to generate that image.
Affected generators:
flux-realism
sd-3.5
sdxl-realvis-v5
sd-1.5-dreamshaper
sd-1.5-epicdream
This affects approximately 19.77% of synthetic images. It was first reported in
discussion… See the full description on the dataset page: https://huggingface.co/datasets/ComplexDataLab/OpenFake.stanford_cars
Stanford Cars Dataset
Dataset Overview
Splits:
Training: 8144 images used for model training.
Test: 8041 images used for evaluation.
Contrast: 8041 images with high contrast for robustness testing.
Gaussian Noise: 8041 images corrupted by Gaussian noise for robustness testing.
Impulse Noise: 8041 images corrupted by impulse noise for robustness testing.
JPEG Compression: 8041 compressed images for robustness testing.
Motion Blur: 8041 images with motion blur for… See the full description on the dataset page: https://huggingface.co/datasets/tanganke/stanford_cars.TreeOfLife-200M
Dataset Card for TreeOfLife-200M
If you are looking for the original release TreeOfLife-200M dataset, as used in training BioCLIP 2 and presented the paper, please see Revision a8f38b4. The dataset, as presented here, was used to train BioCLIP 2.5 Huge; it completes the dataset cleaning process and resolves an issue where Observation.org occurrences were not included in the training data.
With 233 million images representing 933,798 taxa across the tree of life, TreeOfLife-200M… See the full description on the dataset page: https://huggingface.co/datasets/imageomics/TreeOfLife-200M.fish-vista
Dataset Card for Fish-Visual Trait Analysis (Fish-Vista)
Note that the '</Use this dataset>' option will only load the CSV files. To download the entire dataset, including all processed images and segmentation annotations, refer to Instructions for downloading dataset and images.
See Example Code to Use the Segmentation Dataset
Figure 1. A schematic representation of the different tasks in Fish-Vista Dataset.
Instructions for downloading dataset… See the full description on the dataset page: https://huggingface.co/datasets/imageomics/fish-vista.game_character_skins
Game Character Skins Dataset
Summary
This comprehensive dataset contains game character skins and artwork from multiple popular mobile and PC games, providing a rich collection of character visual assets for computer vision research and game development applications. The dataset spans eight major game titles including Arknights, Azur Lane, Blue Archive, Fate/Grand Order, Genshin Impact, Girls' Frontline, Neural Cloud, Nikke, Path to Nowhere, and Honkai: Star Rail… See the full description on the dataset page: https://huggingface.co/datasets/deepghs/game_character_skins.svhn
Dataset Card for Street View House Numbers
Dataset Summary
SVHN is a real-world image dataset for developing machine learning and object recognition algorithms with minimal requirement on data preprocessing and formatting. It can be seen as similar in flavor to MNIST (e.g., the images are of small cropped digits), but incorporates an order of magnitude more labeled data (over 600,000 digit images) and comes from a significantly harder, unsolved, real world problem… See the full description on the dataset page: https://huggingface.co/datasets/ufldl-stanford/svhn.imagenet1k-256-wdsThis is imagenet1k in webdataset format. Images are stored as jpg files. Every image has been resized to a maximum side length of 256. That means that if an image in the original dataset was 1000 by 500, the new size will be 256 by 128. Images with a maximum side length of under 256 were not resized.
The total size of all dataset files is 57.8 GB, there are 1,281,167 rows in the training split and 50,000 rows in the validation split.
sun397
SUN397 dataset
The database contains 397 categories subset from the SUN dataset for Scene Recognition used in the following paper.
The number of images varies across categories, but there are at least 100 images per category, and 108,754 images in total.
All images are in jpg format. The images provided here are for research purposes only.
The file ClassName.txt contains the name list for the 397 categories.
Please cite the following paper if you use this dataset in your research.… See the full description on the dataset page: https://huggingface.co/datasets/tanganke/sun397.imagenet_1k_resized_256
Dataset Card for "imagenet_1k_resized_256"
Dataset summary
The same ImageNet dataset but all the smaller side resized to 256.
A lot of pretraining workflows contain resizing images to 256 and random cropping to 224x224, this is why 256 is chosen.
The resized dataset can also be downloaded much faster and consume less space than the original one.
See here for detailed readme.
Dataset Structure
Below is the example of one row of data. Note that the labels in… See the full description on the dataset page: https://huggingface.co/datasets/evanarlian/imagenet_1k_resized_256.fashion_mnist
Dataset Card for FashionMNIST
Dataset Summary
Fashion-MNIST is a dataset of Zalando's article images—consisting of a training set of 60,000 examples and a test set of 10,000 examples. Each example is a 28x28 grayscale image, associated with a label from 10 classes. We intend Fashion-MNIST to serve as a direct drop-in replacement for the original MNIST dataset for benchmarking machine learning algorithms. It shares the same image size and structure of training and testing… See the full description on the dataset page: https://huggingface.co/datasets/zalando-datasets/fashion_mnist.oxford-iiit-pet
The Oxford-IIIT Pet Dataset
Description
A 37 category pet dataset with roughly 200 images for each class. The images have a large variations in scale, pose and lighting.
This instance of the dataset uses standard label ordering and includes the standard train/test splits. Trimaps and bbox are not included, but there is an image_id field that can be used to reference those annotations from official metadata.
Website: https://www.robots.ox.ac.uk/~vgg/data/pets/… See the full description on the dataset page: https://huggingface.co/datasets/timm/oxford-iiit-pet.femnist
Dataset Card for FEMNIST
The FEMNIST dataset is a part of the LEAF benchmark.
It represents image classification of handwritten digits, lower and uppercase letters, giving 62 unique labels.
Dataset Details
Dataset Description
Each sample is comprised of a (28x28) grayscale image, writer_id, hsf_id, and character.
Curated by: LEAF
License: BSD 2-Clause License
Dataset Sources
The FEMNIST is a preprocessed (in a way that resembles preprocessing for… See the full description on the dataset page: https://huggingface.co/datasets/flwrlabs/femnist.GroMo25
GroMo25: Multiview Time-Series Plant Image Dataset for Age Estimation and Leaf Counting
Dataset Summary
GroMo25 is a multiview, time-series plant image dataset designed for plant age estimation (in days) and leaf counting tasks in precision agriculture. It contains high-quality images of four crop species — Wheat, Okra, Radish, and Mustard — captured over multiple days under controlled conditions. Each plant is photographed from 24 angles across 5 vertical levels per day… See the full description on the dataset page: https://huggingface.co/datasets/MrigLabIITRopar/GroMo25.MIC21
[!NOTE]
Dataset origin: https://live.european-language-grid.eu/catalogue/corpus/18029/
MIC21
Original description
One of the processing tasks for large multimodal data streams is automatic image description (image classification, object segmentation and classification). Although the number and the diversity of image datasets is constantly expanding, still there is a huge demand for more datasets in terms of variety of domains and object classes covered. The goal of the… See the full description on the dataset page: https://huggingface.co/datasets/FrancophonIA/MIC21.FSC147Yet another mirror for FSC147 counting dataset
All credits go to https://github.com/cvlab-stonybrook/LearningToCountEverything
eurosat
Dataset Card for EuroSAT
Dataset Source
Paper with code
Usage
from datasets import load_dataset
dataset = load_dataset('tranganke/eurosat')
Data Fields
The dataset contains the following fields:
image: An image in RGB format.
label: The label for the image, which is one of 10 classes:
0: annual crop land
1: forest
2: brushland or shrubland
3: highway or road
4: industrial buildings or commercial buildings
5: pasture land
6: permanent crop land… See the full description on the dataset page: https://huggingface.co/datasets/tanganke/eurosat.gtsrb
Dataset Card for German Traffic Sign Recognition Benchmark
This dataset contains images of 43 classes of traffic signs. It is intended for developing and benchmarking traffic sign recognition systems.
Dataset Details
Dataset Description
The German Traffic Sign Recognition Benchmark (GTSRB) is a multi-class classification dataset featuring 43 classes of traffic signs.
The images were cropped from a larger set of images to focus on the traffic sign and eliminate… See the full description on the dataset page: https://huggingface.co/datasets/tanganke/gtsrb.svg-benchmark
Rapidata Static SVG Generation Benchmark
Built by Rapidata.
This dataset contains 1,918,367 human responses, collected with the
Rapidata Python SDK, comparing how well 42 frontier LLMs generate
static SVGs from text prompts. Each row is a head-to-head comparison between two models' renders of
the same prompt, scored by human annotators on one of three questions (Preference, Coherence, Alignment).
The SVGs are produced as raw <svg> markup by the models, rasterized to 768×768 PNGs… See the full description on the dataset page: https://huggingface.co/datasets/Rapidata/svg-benchmark.