Dataset Description

A mini version of ImageNet-1k with 100 of 1000 classes present. Unlike some 'mini' variants this one includes the original images at their original sizes. Many such subsets downsample to 84x84 or other smaller resolutions.

  Data Splits





  Train

50000 samples from ImageNet-1k train split

  Validation

10000 samples from ImageNet-1k train split

  Test

5000 samples from ImageNet-1k validation split (all 50 samples per class)… See the full description on the dataset page: https://huggingface.co/datasets/timm/mini-imagenet.

dnth/mini-imagenet dataset hosted on Hugging Face and contributed by the HF Datasets community

Dataset

This dataset was created by Steven Leo

Released under Data files © Original Authors

Contents

The miniImagenet and tieredImageNet datasets are used for few-shot classification experiments.

The mini-ImageNet dataset is a subset of the ImageNet dataset, containing 60,000 images from 100 classes.

Dataset

This dataset was created by Arjun Ashok

Contents

It contains the following files:

MiniImageNet pickle files for learn2learn library - please move these files to ~/data/l2l_data (or whereever your l2l_data directory is) in order for MAML and USL experiments to properly run and recognize MiniImageNet.

The mini-ImageNet dataset was proposed by Vinyals et al. for few-shot learning evaluation. Its complexity is high due to the use of ImageNet images but requires fewer resources and infrastructure than running on the full ImageNet dataset. In total, there are 100 classes with 600 samples of 84×84 color images per class. These 100 classes are divided into 64, 16, and 20 classes respectively for sampling tasks for meta-training, meta-validation, and meta-test.

The dataset used in the paper is a mini-imagenet dataset, with 10100 instances, 21168 attributes, and 100 categories. The dataset is used for clustering and evaluation of the proposed method.

yonigo/mini-imagenet-clip-embeddings dataset hosted on Hugging Face and contributed by the HF Datasets community

Average test set classification accuracy on miniimageNet.

miniImageNet, tieredImageNet, Caltech-USCD birds-200-2011

Dataset Card for "mini_imagenet"

More Information needed

Dataset

This dataset was created by Hylan-J

Released under MIT

Contents

Controlled Noisy Web Labels is a collection of ~212,000 URLs to images in which every image is carefully annotated by 3-5 labeling professionals by Google Cloud Data Labeling Service. Using these annotations, it establishes the first benchmark of controlled real-world label noise from the web.

We provide the Red Mini-ImageNet (real-world web noise) and Blue Mini-ImageNet configs: - controlled_noisy_web_labels/mini_imagenet_red - controlled_noisy_web_labels/mini_imagenet_blue

Each config contains ten variants with ten noise-levels p from 0% to 80%. The validation set has clean labels and is shared across all noisy training sets. Therefore, each config has the following splits:

• train_00
• train_05
• train_10
• train_15
• train_20
• train_30
• train_40
• train_50
• train_60
• train_80
• validation

The details for dataset construction and analysis can be found in the paper. All images are resized to 84x84 resolution.

To use this dataset:

import tensorflow_datasets as tfds

ds = tfds.load('controlled_noisy_web_labels', 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/controlled_noisy_web_labels-mini_imagenet_red-1.0.0.png" alt="Visualization" width="500px">

MiniImageNet, Caltech-UCSD Birds 200-2011, TieredImageNet, OfficeHome

MiniImageNet, TieredImageNet, and CUB are used for few-shot learning tasks.

Dataset

This dataset was created by dotran0101

Contents

Deep neural networks (DNNs) are known to perform well when deployed to test distributions that shares high similarity with the training distribution. Feeding DNNs with new data sequentially that were unseen in the training distribution has two major challenges — fast adaptation to new tasks and catastrophic forgetting of old tasks. Such difficulties paved way for the on-going research on few-shot learning and continual learning. To tackle these problems, we introduce Attentive Independent Mechanisms (AIM). We incorporate the idea of learning using fast and slow weights in conjunction with the decoupling of the feature extraction and higher-order conceptual learning of a DNN. AIM is designed for higher-order conceptual learning, modeled by a mixture of experts that compete to learn independent concepts to solve a new task. AIM is a modular component that can be inserted into existing deep learning frameworks. We demonstrate its capability for few-shot learning by adding it to SIB and trained on MiniImageNet and CIFAR-FS, showing significant improvement. AIM is also applied to ANML and OML trained on Omniglot, CIFAR-100 and MiniImageNet to demonstrate its capability in continual learning.

Average test set classification accuracy on Omniglot.