datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
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.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.American-Sign-Language-MNIST
Dataset Card for ASL-MNIST
This is a FiftyOne dataset with 34,627 samples of American Sign Language (ASL) alphabet images, converted from the original Kaggle Sign Language MNIST dataset into a format optimized for computer vision workflows.
Installation
If you haven't already, install FiftyOne:
pip install -U fiftyone
Usage
import fiftyone as fo
from fiftyone.utils.huggingface import load_from_hub
# Load the dataset
# Note: other available arguments… See the full description on the dataset page: https://huggingface.co/datasets/Voxel51/American-Sign-Language-MNIST.MNIST_train
IllusionMNIST — Training Set
Dataset summary
This repository contains the training split of IllusionMNIST, introduced in Illusory VQA: Benchmarking and Enhancing Multimodal Models on Visual Illusions. The dataset is intended for training models to recognize MNIST digits embedded as visual illusions (pareidolia) in generated scenes and to reject images that contain no illusion.
MNIST source-condition images were sampled and resized to 512 × 512 pixels, combined… See the full description on the dataset page: https://huggingface.co/datasets/VQA-Illusion/MNIST_train.MNIST_test
IllusionMNIST — Test Set
Dataset summary
This repository contains the public test split of IllusionMNIST, introduced in Illusory VQA: Benchmarking and Enhancing Multimodal Models on Visual Illusions. Every indexed example can be compared across source-condition, illusion, filtered-illusion, illusionless-control, and filtered-illusionless-control images.
The source-condition images are sampled from MNIST and resized to 512 × 512 pixels. Illusion images were… See the full description on the dataset page: https://huggingface.co/datasets/VQA-Illusion/MNIST_test.curated-mnist
Dataset Card for CuratedMNIST
This dataset is a curated version of the classic MNIST dataset, enriched with model predictions, embeddings, and various analytical fields generated using the FiftyOne library. It was created as part of the "Image Classification and Dataset Curation with FiftyOne and PyTorch" tutorial to demonstrate practical computer vision workflows.
The dataset contains the original 60,000 training and 10,000 test samples, with the training set further split into… See the full description on the dataset page: https://huggingface.co/datasets/Voxel51/curated-mnist.MNIST
🖼️ MNIST (Extracted from PyTorch Vision)
MNIST is a classic dataset of handwritten digits, widely used for image classification tasks in machine learning.
ℹ️ Dataset Details
📖 Dataset Description
The MNIST database of handwritten digits is a commonly used benchmark dataset in machine learning. It consists of 70,000 grayscale images of handwritten digits (0-9), each with a size of 28x28 pixels. The dataset is split into 60,000 training images and 10,000… See the full description on the dataset page: https://huggingface.co/datasets/p2pfl/MNIST.mnist-cleaned-full
Dataset Card for 2025.11.21.16.40.44.970939
This is a FiftyOne dataset with 69807 samples.
Installation
If you haven't already, install FiftyOne:
pip install -U fiftyone
Usage
import fiftyone as fo
from fiftyone.utils.huggingface import load_from_hub
# Load the dataset
# Note: other available arguments include 'max_samples', etc
dataset = load_from_hub("Linus-L/mnist-cleaned-full")
# Launch the App
session = fo.launch_app(dataset)
Dataset… See the full description on the dataset page: https://huggingface.co/datasets/Linus-L/mnist-cleaned-full.tactile-mnist-touch-syn-single-t32-320x240Documentation is available at https://github.com/TimSchneider42/tactile-mnist/blob/main/doc/datasets.md#touch-datasets.
colored_mnist_28
Colored MNIST Dataset
A comprehensive dataset of MNIST digits with RGB colored backgrounds, designed for multi-objective classification tasks including digit recognition and background color classification.
Overview
This dataset extends the classic MNIST digit dataset by adding colored backgrounds (blue, green, red) to create a multi-label classification challenge. Each image contains:
Digit label: 0-9 (digit recognition)
Color label: blue, green, red (background… See the full description on the dataset page: https://huggingface.co/datasets/FrankCCCCC/colored_mnist_28.tactile-mnist-touch-starstruck-syn-single-t32-320x240Documentation is available at https://github.com/TimSchneider42/tactile-mnist/blob/main/doc/datasets.md#touch-datasets.
mnist-lance
MNIST (Lance Format)
A Lance-formatted version of the classic MNIST handwritten-digit dataset covering 70,000 28×28 grayscale digits across ten balanced classes. Each row carries inline PNG bytes, the digit label, the human-readable class name, and a cosine-normalized CLIP image embedding, all backed by a bundled IVF_PQ vector index plus scalar indices on the label columns and available directly from the Hub at hf://datasets/lance-format/mnist-lance/data.
Key features… See the full description on the dataset page: https://huggingface.co/datasets/lance-format/mnist-lance.fashion-mnist-lance
Fashion-MNIST (Lance Format)
A Lance-formatted version of Fashion-MNIST covering 70,000 28×28 grayscale clothing images across ten balanced apparel classes. Each row carries inline PNG bytes, the integer label, the human-readable class name, and a cosine-normalized CLIP image embedding, all backed by a bundled IVF_PQ vector index plus scalar indices on the label columns and available directly from the Hub at hf://datasets/lance-format/fashion-mnist-lance/data.
Key… See the full description on the dataset page: https://huggingface.co/datasets/lance-format/fashion-mnist-lance.American-Sign-Language-MNIST
Dataset Card for ASL-MNIST
This is a FiftyOne dataset with 34,627 samples of American Sign Language (ASL) alphabet images, converted from the original Kaggle Sign Language MNIST dataset into a format optimized for computer vision workflows.
Installation
If you haven't already, install FiftyOne:
pip install -U fiftyone
Usage
import fiftyone as fo
from fiftyone.utils.huggingface import load_from_hub
# Load the dataset
# Note: other available… See the full description on the dataset page: https://huggingface.co/datasets/saifmughal23/American-Sign-Language-MNIST.mnist-100
MNIST-100
A tiny subset of the MNIST training set:
10 images per digit (0-9), 100 images in total.
Source: torchvision.datasets.MNIST (train split)
Sampling: random, seed 42
Format: 28x28 grayscale PNG, ImageFolder layout (train/<label>/<file>.png)
File names contain the index in the original MNIST training set
Usage
from datasets import load_dataset
ds = load_dataset("MichielBontenbal/mnist-100")
print(ds["train"][0]) # {'image': <PIL.Image>, 'label': 0}
quickdraw-mnist
QuickDraw-MNIST
QuickDraw-MNIST is a 20-class sketch-recognition dataset prepared for Texas A&M's CSCE 624 (Sketch Recognition) class.
The data is sourced from Google's Quick, Draw! dataset.
Dataset Structure
Number of images: 100,000
Number of classes: 20
Images: 64 x 64 grayscale
Labels: integer class ids with a human-readable label_name column
Classes: The Eiffel Tower, airplane, angel, bed, chair, clock, diamond, donut, fork, frog, hourglass, leaf, line, mushroom… See the full description on the dataset page: https://huggingface.co/datasets/oriyonay/quickdraw-mnist.mnist-cleaned-up
Dataset Card for cleaned-up-mnist-training-set
This is a FiftyOne dataset with 505 samples.
Installation
If you haven't already, install FiftyOne:
pip install -U fiftyone
Usage
import fiftyone as fo
from fiftyone.utils.huggingface import load_from_hub
# Load the dataset
# Note: other available arguments include 'max_samples', etc
dataset = load_from_hub("Linus-L/mnist-cleaned-up")
# Launch the App
session = fo.launch_app(dataset)
Dataset… See the full description on the dataset page: https://huggingface.co/datasets/Linus-L/mnist-cleaned-up.curated_mnist
Dataset Card for mnist-curated
This is a FiftyOne dataset with 70000 samples.
Installation
If you haven't already, install FiftyOne:
pip install -U fiftyone
Usage
import fiftyone as fo
from fiftyone.utils.huggingface import load_from_hub
# Load the dataset
# Note: other available arguments include 'max_samples', etc
dataset = load_from_hub("maxspeer/curated-mnist5")
# Launch the App
session = fo.launch_app(dataset)
Dataset Details
This… See the full description on the dataset page: https://huggingface.co/datasets/maxspeer/curated_mnist.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/mteb/mnist.A-MNISTThe dataset is built on top of MNIST.
It consists from 130K of images in 10 classes - 120K training and 10K test samples.
The training set was augmented with additional 60K images.ufo-mnist
UFO-MNIST
Table of Contents
Why UFO-MNIST
Get the Data
Labels
Usage
Benchmark
Visualization
Contributing
Citing UFO-MNIST
License
UFO-MNIST is a dataset of UFO-like spotting patterns and common aerial
lookalikes, consisting of a training set of 8,000 examples and a test set of
2,000 examples. Each example is a 28x28 grayscale image associated with one of
10 labels.
UFO-MNIST is designed as a compact, MNIST-style benchmark for machine-learning
examples and image classifiers. It… See the full description on the dataset page: https://huggingface.co/datasets/tentime/ufo-mnist.curated-mnist
Dataset Card for curated-mnist
This is a FiftyOne dataset with 70000 samples.
Installation
If you haven't already, install FiftyOne:
pip install -U fiftyone
Usage
import fiftyone as fo
from fiftyone.utils.huggingface import load_from_hub
# Load the dataset
# Note: other available arguments include 'max_samples', etc
dataset = load_from_hub("CarloColumbo/curated-mnist")
# Launch the App
session = fo.launch_app(dataset)
Dataset Details… See the full description on the dataset page: https://huggingface.co/datasets/CarloColumbo/curated-mnist.curated-mnist
Dataset Card for curated-mnist
This is a FiftyOne dataset with 70000 samples.
Installation
If you haven't already, install FiftyOne:
pip install -U fiftyone
Usage
import fiftyone as fo
from fiftyone.utils.huggingface import load_from_hub
# Load the dataset
# Note: other available arguments include 'max_samples', etc
dataset = load_from_hub("MatthiasCr/curated-mnist")
# Launch the App
session = fo.launch_app(dataset)
Notebook… See the full description on the dataset page: https://huggingface.co/datasets/MatthiasCr/curated-mnist.tactile-mnist-touch-real-single-t256-320x240Documentation is available at https://github.com/TimSchneider42/tactile-mnist/blob/main/doc/datasets.md#touch-datasets.
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/anonyme449/fashion_mnist.fashion_mnist_ambiguousThe images were created such that they have an unclear ground truth,
i.e., such that they are similar to multiple - but not all - of the datasets classes.
Robust and uncertainty-aware models should be able to detect and flag these ambiguous images.
As such, the dataset should be merged / mixed with the original dataset and we
provide such 'mixed' splits for convenience. Please refer to the dataset card for details.fashion-mnist-mini
AnnotateIt · Open the app · Models & datasets · Documentation
AnnotateIt Fashion-MNIST Mini
Small, deterministic, AnnotateIt-compatible samples derived from Fashion-MNIST.
Upstream revision: b2617bb6d3ffa2e429640350f613e3291e10b141Upstream license: MIT
File
Sample
Tasks
Format
Images
Size
SHA-256
Derived task
fashion-mnist-object-detection-mini.zip
Fashion-MNIST Detection Mini
Detection
COCO
80
0.08 MiB… See the full description on the dataset page: https://huggingface.co/datasets/AnnotateIt/fashion-mnist-mini.mnist_bakery_data
🧁 MNIST Bakery Dataset
A procedurally synthesized variant of the classic MNIST dataset, created using SideFX Houdini and designed for experimentation in data augmentation, synthetic data generation, and model robustness research.
See the ML-Research repository on GitHub for Python notebooks, experiments and the Houdini scene files.
🎯 Purpose
This dataset demonstrates how procedural generation pipelines in 3D tools like Houdini can be used to create high-quality… See the full description on the dataset page: https://huggingface.co/datasets/Arkaen-AtC/mnist_bakery_data.tactile-mnist-touch-real-single-t256-320x240Documentation is available at https://github.com/[REDACTED]/tactile-mnist/blob/main/doc/datasets.md#touch-datasets.
mnist_ambiguousThe images were created such that they have an unclear ground truth,
i.e., such that they are similar to multiple - but not all - of the datasets classes.
Robust and uncertainty-aware models should be able to detect and flag these ambiguous images.
As such, the dataset should be merged / mixed with the original dataset and we
provide such 'mixed' splits for convenience. Please refer to the dataset card for details.