525 datasets found

Hyperspectral images classification is a high technical remote sensing software. The purpose is to reproduce a thematic map. The HSI contains more than a hundred hyperspectral...

The MNIST-Superpixel dataset is a collection of images with superpixels, where each superpixel is a region of similar color.

A dataset of 50 animal species with 85 attribute annotations, used for AWA2 dataset.

A dataset of 14,340 images with 102 attributes, used for SUN Attribute (SUN) dataset.

A dataset of 11,788 images of 200 bird species, used for Caltech-UCSD Birds-200-2011 (CUB) dataset.

A dataset of 15,339 images with 64 attributes, used for Attribute Pascal and Yahoo (APY) dataset.

A dataset for active semi-supervised learning for class imbalanced datasets

A dataset for active semi-supervised learning for class imbalanced datasets

The Taskonomy dataset contains 16 types of annotation, including object classification, scene classification, depth estimation, surface normals, curvature estimation, edge...

The MNIST-Fellowship dataset is composed of three datasets (MNIST, Fashion-MNIST, and KMNIST) each composed of 10 classes.

Zero-shot models suffer due to artifacts inherited from pretraining. A particularly detrimental artifact, caused by unbalanced web-scale pretraining data, is mismatched label...

The dataset used in the paper is CIFAR-10 and CIFAR-100, as well as SVHN.

Correlation responses obtained by correlation filters for image classification

UT Zappos50K dataset

A variant of CIFAR-100 dataset, where the original dataset is split into 20 disjoint tasks, each consisting of 2,500 samples from 5 classes.

The dataset used in the paper is FashionMNIST and CIFAR-10, which are commonly used datasets for image classification tasks.

The dataset used in this study for image and text classification tasks.

A dataset of 3000 images collected on a public roadway for front seat vehicle occupancy detection.

Two datasets used for multi-task learning, CIFAR-100 and AGNews.

The dataset used in the paper is the ImageNet noise dataset, which contains 60,000 32x32 color images with random labels.