datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
imagenet-r
ImageNet-R
This repo is made to facilitate the evaluation of various pretraining models. It's constructed from the source file provided by official implementation.
Usage
from datasets import load_dataset
dataset = load_dataset('axiong/imagenet-r')
Dataset Summary
ImageNet-R(endition) contains art, cartoons, deviantart, graffiti, embroidery, graphics, origami, paintings, patterns, plastic objects, plush objects, sculptures, sketches, tattoos, toys, and video… See the full description on the dataset page: https://huggingface.co/datasets/axiong/imagenet-r.Imagenet21KNOTE: I have recaptioned all images here
This dataset is the entire 21K ImageNet dataset with about 13 million examples and about 19 thousand classes as strings
(for some reason it only had ~19K classes instead of 21K).
The images are in PNG format. They can be decoded like in the following example
import io
from PIL import Image
Image.open(io.BytesIO(row["image"]))
where row["image"] are the raw image bytes.
imagenet-1k-vl-enriched
Visualize on Visual Layer
Imagenet-1K-VL-Enriched
An enriched version of the ImageNet-1K Dataset with image caption, bounding boxes, and label issues!
With this additional information, the ImageNet-1K dataset can be extended to various tasks such as image retrieval or visual question answering.
The label issues helps to curate a cleaner and leaner dataset.
Description
The dataset consists of 6 columns:
image_id: The original filename of the image from… See the full description on the dataset page: https://huggingface.co/datasets/visual-layer/imagenet-1k-vl-enriched.Imagenet21K_RecaptionThis dataset is the entire 21K ImageNet dataset with about 13 million examples and about 19 thousand classes as strings
(for some reason it only had ~19K classes instead of 21K).
If you want an even larger set of images, I have a recaptioned CC12M and ImageNet dataset: https://huggingface.co/datasets/gmongaras/CC12M_and_Imagenet21K_Recap
The images are in PNG format. They can be decoded like in the following example
import io
from PIL import Image
Image.open(io.BytesIO(row["image"]))
where… See the full description on the dataset page: https://huggingface.co/datasets/gmongaras/Imagenet21K_Recaption.features-dinov3-vith16plus-224-imagenet-22k-wdsCC12M_and_Imagenet21K_RecapI removed all low quality data and uploaded it here
This dataset is the entire 21K ImageNet dataset with about 13 million examples and about 19 thousand classes as strings
(for some reason it only had ~19K classes instead of 21K) as well as the entire CC12M dataset, recaptioned.
If you just want the recaptioned Imagenet dataset, I have that here
I obtained the CC12M form others. CC12M is a dataset with 12 million images created in 2021. Unfortunately the downloader provided by Google has many… See the full description on the dataset page: https://huggingface.co/datasets/gmongaras/CC12M_and_Imagenet21K_Recap.ImageNet100
Dataset Card for Dataset Name
This dataset card aims to be a base template for new datasets. It has been generated using this raw template.
Dataset Details
This is Huggingface dataset version of https://www.kaggle.com/datasets/ambityga/imagenet100. All credits are given to the original author and please cite the original author.
Acknowledgements
Olga Russakovsky, Jia Deng, Hao Su, Jonathan Krause, Sanjeev Satheesh, Sean Ma, Zhiheng Huang, Andrej Karpathy… See the full description on the dataset page: https://huggingface.co/datasets/ilee0022/ImageNet100.imagenet-1k-rand_blur
Dataset Card for "imagenet-1k-rand_blur"
More Information needed
imagenet-hard-4K
Dataset Card for "Imagenet-Hard-4K"
Project Page - Paper - Github
ImageNet-Hard-4K is 4K version of the original ImageNet-Hard dataset, which is a new benchmark that comprises 10,980 images collected from various existing ImageNet-scale benchmarks (ImageNet, ImageNet-V2, ImageNet-Sketch, ImageNet-C, ImageNet-R, ImageNet-ReaL, ImageNet-A, and ObjectNet). This dataset poses a significant challenge to state-of-the-art vision models as merely zooming in often fails to improve their… See the full description on the dataset page: https://huggingface.co/datasets/taesiri/imagenet-hard-4K.imagenet-dog-15PUG_ImageNet
PUG: ImageNet
The PUG: ImageNet dataset contains 88,328 pre-rendered images based on Unreal Engine using 724 assets representing 151 ImageNet classes with 64 environments, 7 sizes, 9 textures, 18 different camera orientations, 18 different character orientations and 7 light intensities. In contrast to PUG: Animals, PUG: ImageNet was created by varying only a single factor at a time (which explains the lower number of images than PUG: Animals despite using more factors). The main… See the full description on the dataset page: https://huggingface.co/datasets/facebook/PUG_ImageNet.semasia-imagenet-1k
Latents for imagenet-1k (timm)
This repository hosts precomputed latent representations (embeddings) extracted from timm image-classification backbones on imagenet-1k, released as part of SEMASIA — a large-scale resource for studying semantic communication, cross-model latent space alignment, and explainability.
Each config corresponds to a single model;
only that model's Parquet files are read on load_dataset.
Usage
Load with datasets… See the full description on the dataset page: https://huggingface.co/datasets/spaicom-lab/semasia-imagenet-1k.ImageNet1K-T2I-QwenVL-QwenImageimagenet_288_webpA duplicate of the webdataset version of ImageNet-1k, except each image is resized so that the shorter side is 288 pixels and each image is WEBP compressed using a quality level of 85.
The compression ratio is roughly 5:1 compared to JPEG (~160 GB) and 25:1 compared to raw pixel storage (~800GB).
imagenet-w21-p
Dataset Summary
This is a subset of the full Winter21, filtered according to https://github.com/Alibaba-MIIL/ImageNet21K. This instance contains 10450 classes with a train and validation split.
Processing
I performed some processing while sharding this dataset:
Synsets were filtered according to ImageNet-21-P scripts
Images were re-encoded in WEBP
Additional Information
Dataset Curators
Authors of [1] and [2]:
Olga Russakovsky
Jia Deng
Hao Su… See the full description on the dataset page: https://huggingface.co/datasets/timm/imagenet-w21-p.resnet50-imagenet-embeddingsimagenet-1k-rand_colorjitterimagenet1k_invae-latents_dinov2_pcaImagenet-Classimagenet1k_val_3kimagenet-aimagenet-1k-256The dataset contains imagenet-1k resized to 256x256, along with corresponding captions generated by gemini-2.0-flash, and discrete tokens generated by NVIDIA Cosmos tokenizer.
Specifically the following steps are done in order
Start with hf:timm/imagenet-1k-wds
Generate captions for the images with gemini-2.0-flash with the following config:.
_SYSTEM_PROMPT = '''Caption the ImageNet-21K in a single, detailed paragraph, without introductory phrases like "This image showcases", "Here's a… See the full description on the dataset page: https://huggingface.co/datasets/pshishodia/imagenet-1k-256.Imagenet1k_sample_validation
Dataset Card for "Imagenet1k_sample_validation"
More Information needed
imagenet-1k-rand_canny_colorgrid
Dataset Card for "imagenet-1k-rand_canny_colorgrid"
More Information needed
imagenet-r-test
imagenet-r-test-vqa
This dataset has been converted to VQA format with:
problem: Question with multiple-choice options (format: "What type of object is in the photo?\n Please choose one from list [ ... ].")
solution: Answer in <answer>classname</answer> format
split: 'test'
Conversion Details
Original dataset: imagenet-r-test
Total samples: 30000
Number of classes: 200
Choices per question: 80
Format: Parquet (consistent with original dataset)
Usage
from… See the full description on the dataset page: https://huggingface.co/datasets/kaze-desu/imagenet-r-test.ImageNet1K-T2I-QwenVL-FLUXimagenet_sample20
Introduction
This is a tiny subset of ImageNet (imagenet-w21-wds). For each label, there are 20 sampled images. Those labels that have less than 20 images are skipped.
Citation
Please cite the paper if you use this dataset.
@article{imagenet15russakovsky,
Author = {Olga Russakovsky and Jia Deng and Hao Su and Jonathan Krause and Sanjeev Satheesh and Sean Ma and Zhiheng Huang and Andrej Karpathy and Aditya Khosla and Michael Bernstein and Alexander C. Berg and Li… See the full description on the dataset page: https://huggingface.co/datasets/yyyyifan/imagenet_sample20.Imagenet_Sketchimagenet-1k-rand_entropyV3Det_ImageNet21k_Cls_100
Install datasets package
First, make sure you have the datasets library installed. If not, you can install it using:
pip install datasets
Load Dataset from Arrow File
Download all arrow files to local_path. The follow is how to load arrow files and decode image:
from datasets import load_from_disk
from io import BytesIO
import base64
from PIL import Image
import mmengine
# Path to your Arrow dataset directory
arrow_dataset_path = 'path_to_your_arrow_dataset_directory'… See the full description on the dataset page: https://huggingface.co/datasets/yhcao/V3Det_ImageNet21k_Cls_100.