The ImageNet dataset contains 14,197,122 annotated images according to the WordNet hierarchy. Since 2010 the dataset is used in the ImageNet Large Scale Visual Recognition Challenge (ILSVRC), a benchmark in image classification and object detection. The publicly released dataset contains a set of manually annotated training images. A set of test images is also released, with the manual annotations withheld. ILSVRC annotations fall into one of two categories: (1) image-level annotation of a binary label for the presence or absence of an object class in the image, e.g., “there are cars in this image” but “there are no tigers,” and (2) object-level annotation of a tight bounding box and class label around an object instance in the image, e.g., “there is a screwdriver centered at position (20,25) with width of 50 pixels and height of 30 pixels”. The ImageNet project does not own the copyright of the images, therefore only thumbnails and URLs of images are provided.

Total number of non-empty WordNet synsets: 21841 Total number of images: 14197122 Number of images with bounding box annotations: 1,034,908 Number of synsets with SIFT features: 1000 Number of images with SIFT features: 1.2 million

ImageNet-R is a set of images labelled with ImageNet labels that were obtained by collecting art, cartoons, deviantart, graffiti, embroidery, graphics, origami, paintings, patterns, plastic objects, plush objects, sculptures, sketches, tattoos, toys, and video game renditions of ImageNet classes. ImageNet-R has renditions of 200 ImageNet classes resulting in 30,000 images. by collecting new data and keeping only those images that ResNet-50 models fail to correctly classify. For more details please refer to the paper.

The label space is the same as that of ImageNet2012. Each example is represented as a dictionary with the following keys:

• 'image': The image, a (H, W, 3)-tensor.
• 'label': An integer in the range [0, 1000).
• 'file_name': A unique sting identifying the example within the dataset.

To use this dataset:

import tensorflow_datasets as tfds

ds = tfds.load('imagenet_r', split='train')
for ex in ds.take(4):
 print(ex)

See the guide for more informations on tensorflow_datasets.

https://storage.googleapis.com/tfds-data/visualization/fig/imagenet_r-0.2.0.png" alt="Visualization" width="500px">

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+). In ImageNet, we aim to provide on average 1000 images to illustrate each synset. Images of each concept are quality-controlled and human-annotated. In its completion, we hope ImageNet will offer tens of millions of cleanly sorted images for most of the concepts in the WordNet hierarchy.

The test split contains 100K images but no labels because no labels have been publicly released. We provide support for the test split from 2012 with the minor patch released on October 10, 2019. In order to manually download this data, a user must perform the following operations:

1. Download the 2012 test split available here.
2. Download the October 10, 2019 patch. There is a Google Drive link to the patch provided on the same page.
3. Combine the two tar-balls, manually overwriting any images in the original archive with images from the patch. According to the instructions on image-net.org, this procedure overwrites just a few images.

The resulting tar-ball may then be processed by TFDS.

To assess the accuracy of a model on the ImageNet test split, one must run inference on all images in the split, export those results to a text file that must be uploaded to the ImageNet evaluation server. The maintainers of the ImageNet evaluation server permits a single user to submit up to 2 submissions per week in order to prevent overfitting.

To evaluate the accuracy on the test split, one must first create an account at image-net.org. This account must be approved by the site administrator. After the account is created, one can submit the results to the test server at https://image-net.org/challenges/LSVRC/eval_server.php The submission consists of several ASCII text files corresponding to multiple tasks. The task of interest is "Classification submission (top-5 cls error)". A sample of an exported text file looks like the following:

771 778 794 387 650
363 691 764 923 427
737 369 430 531 124
755 930 755 59 168

The export format is described in full in "readme.txt" within the 2013 development kit available here: https://image-net.org/data/ILSVRC/2013/ILSVRC2013_devkit.tgz Please see the section entitled "3.3 CLS-LOC submission format". Briefly, the format of the text file is 100,000 lines corresponding to each image in the test split. Each line of integers correspond to the rank-ordered, top 5 predictions for each test image. The integers are 1-indexed corresponding to the line number in the corresponding labels file. See labels.txt.

To use this dataset:

import tensorflow_datasets as tfds

ds = tfds.load('imagenet2012_subset', split='train')
for ex in ds.take(4):
 print(ex)

See the guide for more informations on tensorflow_datasets.

https://storage.googleapis.com/tfds-data/visualization/fig/imagenet2012_subset-1pct-5.0.0.png" alt="Visualization" width="500px">

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… See the full description on the dataset page: https://huggingface.co/datasets/zh-plus/tiny-imagenet.

This repository contains the mapping from integer id's to actual label names (in HuggingFace Transformers typically called id2label) for several datasets. Current datasets include:

ImageNet-1k ImageNet-22k (also called ImageNet-21k as there are 21,843 classes) COCO detection 2017 COCO panoptic 2017 ADE20k (actually, the MIT Scene Parsing benchmark, which is a subset of ADE20k) Cityscapes VQAv2 Kinetics-700 RVL-CDIP PASCAL VOC Kinetics-400 ...

You can read in a label file as follows (using… See the full description on the dataset page: https://huggingface.co/datasets/huggingface/label-files.

Imagenet32 is a huge dataset made up of small images called the down-sampled version of Imagenet. Imagenet32 is composed of 1,281,167 training data and 50,000 test data with 1,000 labels.

Dataset Summary

This is a copy of the full ImageNet dataset consisting of all of the original 21841 clases. It also contains labels in a separate field for the '12k' subset described at at (https://github.com/rwightman/imagenet-12k, https://huggingface.co/datasets/timm/imagenet-12k-wds) This dataset is from the original fall11 ImageNet release which has been replaced by the winter21 release which removes close to 3000 synsets containing people, a number of these are of an… See the full description on the dataset page: https://huggingface.co/datasets/timm/imagenet-22k-wds.

ImageNet-PI is a relabelled version of the standard ILSVRC2012 ImageNet dataset in which the labels are provided by a collection of 16 deep neural networks with different architectures pre-trained on the standard ILSVRC2012. Specifically, the pre-trained models are downloaded from tf.keras.applications.

In addition to the new labels, ImageNet-PI also provides meta-data about the annotation process in the form of confidences of the models on their labels and additional information about each model.

For more information see: ImageNet-PI

To use this dataset:

import tensorflow_datasets as tfds

ds = tfds.load('imagenet_pi', split='train')
for ex in ds.take(4):
 print(ex)

See the guide for more informations on tensorflow_datasets.

ImageNet-VidVRD dataset contains 1,000 videos selected from ILVSRC2016-VID dataset based on whether the video contains clear visual relations. It is split into 800 training set and 200 test set, and covers common subject/objects of 35 categories and predicates of 132 categories. Ten people contributed to labeling the dataset, which includes object trajectory labeling and relation labeling. Since the ILVSRC2016-VID dataset has the object trajectory annotation for 30 categories already, we supplemented the annotations by labeling the remaining 5 categories. In order to save the labor of relation labeling, we labeled typical segments of the videos in the training set and the whole of the videos in the test set.

Adversarial patches are optimized contiguous pixel blocks in an input image that cause a machine-learning model to misclassify it. However, their optimization is computationally demanding and requires careful hyperparameter tuning. To overcome these issues, we propose ImageNet-Patch, a dataset to benchmark machine-learning models against adversarial patches. It consists of a set of patches optimized to generalize across different models and applied to ImageNet data after preprocessing them with affine transformations. This process enables an approximate yet faster robustness evaluation, leveraging the transferability of adversarial perturbations.

We release our dataset as a set of folders indicating the patch target label (e.g., banana), each containing 1000 subfolders as the ImageNet output classes.

An example showing how to use the dataset is shown below.

code for testing robustness of a model

import os.path

from torchvision import datasets, transforms, models import torch.utils.data

class ImageFolderWithEmptyDirs(datasets.ImageFolder): """ This is required for handling empty folders from the ImageFolder Class. """

def find_classes(self, directory):
  classes = sorted(entry.name for entry in os.scandir(directory) if entry.is_dir())
  if not classes:
    raise FileNotFoundError(f"Couldn't find any class folder in {directory}.")
  class_to_idx = {cls_name: i for i, cls_name in enumerate(classes) if
          len(os.listdir(os.path.join(directory, cls_name))) > 0}
  return classes, class_to_idx

extract and unzip the dataset, then write top folder here

dataset_folder = 'data/ImageNet-Patch'

available_labels = { 487: 'cellular telephone', 513: 'cornet', 546: 'electric guitar', 585: 'hair spray', 804: 'soap dispenser', 806: 'sock', 878: 'typewriter keyboard', 923: 'plate', 954: 'banana', 968: 'cup' }

select folder with specific target

target_label = 954

dataset_folder = os.path.join(dataset_folder, str(target_label)) normalizer = transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]) transforms = transforms.Compose([ transforms.ToTensor(), normalizer ])

dataset = ImageFolderWithEmptyDirs(dataset_folder, transform=transforms) model = models.resnet50(pretrained=True) loader = torch.utils.data.DataLoader(dataset, shuffle=True, batch_size=5) model.eval()

batches = 10 correct, attack_success, total = 0, 0, 0 for batch_idx, (images, labels) in enumerate(loader): if batch_idx == batches: break pred = model(images).argmax(dim=1) correct += (pred == labels).sum() attack_success += sum(pred == target_label) total += pred.shape[0]

accuracy = correct / total attack_sr = attack_success / total

print("Robust Accuracy: ", accuracy) print("Attack Success: ", attack_sr)

Comparison of annotation precision for single category labels using different algorithms.

To enable research on automated alignment/interpretability evaluations, we release the experimental results of our paper "Scale Alone Does not Improve Mechanistic Interpretability in Vision Models" as a separate dataset.

Note that this is the first dataset containing interpretability measurements obtained through psychophysical experiments for multiple explanation methods and models. The dataset contains >120'000 anonymized human responses, each consisting of the final choice, a confidence score, and a reaction time. Out of these >120'000 responses, > 69'000 passed all our quality assertions - this is the main data (see responses_main.csv). The other responses failed (some) quality assertions and might be of lower quality - they should be used with care (see responses_lower_quality.csv). We consider the former the main dataset and provide the latter as data for development/debugging purposes. Furthermore, the dataset contains the used query images as well as the generated explanations for >760 units across nine models.

The dataset itself is a collection of labels and metainformation without the presence of fixed features that should be predictive of a unit's interpretability. Moreover, finding and constructing features that are predictive of the recorded labels will be one of the open challenges posed by this line of research.

Dataset used for paper -> "Rethinking Dataset Compression: Shifting Focus From Labels to Images"

Dataset created according to the paper Squeeze, Recover and Relabel: Dataset Condensation at ImageNet Scale From A New Perspective.

  Basic Usage

from datasets import load_dataset dataset = load_dataset("he-yang/2025-rethinkdc-imagenet-sre2l-ipc-50")

For more information, please refer to the Rethinking-Dataset-Compression

ImageNet-100 is a subset of ImageNet-1k Dataset from ImageNet Large Scale Visual Recognition Challenge 2012. It contains random 100 classes as specified in Labels.json file. Link: https://www.kaggle.com/datasets/ambityga/imagenet100

Experimental results for different K values.

Dataset used for paper -> "Rethinking Dataset Compression: Shifting Focus From Labels to Images"

Dataset created according to the paper Are Large-scale Soft Labels Necessary for Large-scale Dataset Distillation?.

  Basic Usage

from datasets import load_dataset dataset = load_dataset("he-yang/2025-rethinkdc-imagenet-lpld-ipc-10")

For more information, please refer to the Rethinking-Dataset-Compression

Dataset used for paper -> "Rethinking Dataset Compression: Shifting Focus From Labels to Images"

Dataset created according to the paper Deep Learning on a Data Diet: Finding Important Examples Early in Training.

  Basic Usage

from datasets import load_dataset dataset = load_dataset("he-yang/2025-rethinkdc-imagenet-el2n-ipc-100")

For more information, please refer to the Rethinking-Dataset-Compression

Dataset used for paper -> "Rethinking Dataset Compression: Shifting Focus From Labels to Images"

Dataset created according to the paper On the Diversity and Realism of Distilled Dataset: An Efficient Dataset Distillation Paradigm.

  Basic Usage

from datasets import load_dataset dataset = load_dataset("he-yang/2025-rethinkdc-imagenet-rded-ipc-100")

For more information, please refer to the Rethinking-Dataset-Compression

Open Images is a dataset of ~9M images that have been annotated with image-level labels and object bounding boxes.

The training set of V4 contains 14.6M bounding boxes for 600 object classes on 1.74M images, making it the largest existing dataset with object location annotations. The boxes have been largely manually drawn by professional annotators to ensure accuracy and consistency. The images are very diverse and often contain complex scenes with several objects (8.4 per image on average). Moreover, the dataset is annotated with image-level labels spanning thousands of classes.

To use this dataset:

import tensorflow_datasets as tfds

ds = tfds.load('open_images_v4', split='train')
for ex in ds.take(4):
 print(ex)

See the guide for more informations on tensorflow_datasets.

https://storage.googleapis.com/tfds-data/visualization/fig/open_images_v4-original-2.0.0.png" alt="Visualization" width="500px">

The CIFAR-100 dataset (Canadian Institute for Advanced Research, 100 classes) is a subset of the Tiny Images dataset and consists of 60000 32x32 color images. The 100 classes in the CIFAR-100 are grouped into 20 superclasses. There are 600 images per class. Each image comes with a "fine" label (the class to which it belongs) and a "coarse" label (the superclass to which it belongs). There are 500 training images and 100 testing images per class.

The criteria for deciding whether an image belongs to a class were as follows:

The class name should be high on the list of likely answers to the question “What is in this picture?” The image should be photo-realistic. Labelers were instructed to reject line drawings. The image should contain only one prominent instance of the object to which the class refers. The object may be partially occluded or seen from an unusual viewpoint as long as its identity is still clear to the labeler.