Vision

684 Datasets

Datasets


MSR 3D Video

These sequences were used for our video interpolation work described in High-quality video view interpolation using a layered representation, C.L. Zitn…

3d reconstruction, camera, depth, segmentation

Make3D Depth

The Make3D Depth dataset s designed to learn features to estimate scene depth from a single image. This dataset contains aligned image and range data: …

depth estimation, indoor, learning, outdoor, single view

COIL-100

The COIL-100 (Columbia University Image Library) consists of 100 objects. For formal documentation look at the corresponding compressed technical report, …

image classification, image retrieval

Tiny Images

The Tiny Images dataset consists of 79,302,017 images, each being a 32x32 color image. This data is stored in the form of large binary files which can be …

color, image classification, image retrieval, tiny

CIFAR-10 / 100

The CIFAR-10 dataset consists of 60000 32x32 colour images in 10 classes, with 6000 images per class. There are 50000 training images and 10000 test image…

color, image classification, object, patch, scene, tiny

BSDS500

This new dataset is an extension of the BSDS300, where the original 300 images are used for training / validation and 200 fresh images, together with huma…

contour, edge detection, unsupervised segmentation

BSDS300

The goal of this work is to provide an empirical basis for research on image segmentation and boundary detection. To this end, we have collected 12,000 h…

contour, edge detection, unsupervised segmentation

KU Leuven Facade

The KU Leuven Facade dataset is used for architectural styles classification. M. Mathias, A. Martinovic, J. Weissenberg, S. Haegler, L. Van Gool: Automa…

architecture, image classification, procedural reconstruction, urban

USPS Handwritten Digits

Name: Classes Train. Ex. Test. Ex. Features USPS 10 7291 2007 256 8-bit grayscale images of "0" through "9"; handwritten digi…

handwritten, text classification, text recognition

Stroke Width Transform Text

Stroke Width Transform Text dataset is by Boris Epstein and consists of 307 images and XXX text instances. Detecting Text in Natural Scenes with Stroke…

classification, text detection, text recognition