DATA SET
をテンプレートにして作成
[
トップ
] [
新規
|
一覧
|
単語検索
|
最終更新
|
ヘルプ
|
ログイン
]
開始行:
[[MenuBar]]
CENTER:*DATA SET [#o624485f]
***CiFAR-10 [#m923cec7]
The CIFAR-10 is a labeled dataset of 60,000 images of animals and automobiles collected by Alex Krizhevsky, Vinod Nair, and Geoffery Ninton for training object classification models.
-https://www.cs.toronto.edu/~kriz/cifar.html
--Featuers:
---50,000 training images divided into 5 batches of 10,000 images each.
---10,000 testing images of a single batch.
---The images are divided into 10 classes; Airplane, Deer, Cat, Dog, Bird, Frog, Horse, ship, Truck.
---Each image is an RGB image of 32x32x3 pixels.
***CiFAR-100 [#h079723a]
The CIFAR-100 is a labeled dataset of 60,000 images of animals and automobiles collected by Alex Krizhevsky, Vinod Nair, and Geoffery Ninton for training object classification models.
-https://www.cs.toronto.edu/~kriz/cifar.html
--Featuers:
---50,000 training images divided into 5 batches of 10,000 images each.
---10,000 testing images of a single batch.
---The images are divided into 100 classes that are grouped into 20 super classes. Each image is labelled by its class and super class.
---Each image is an RGB image of 32x32x3 pixels.
***MNIST [#b61d656d]
MNIST is a dataset of handwritten digit images. It is commonly used for training and testing machine learning models.
-https://github.com/arXivTimes/arXivTimes/tree/master/datasets
-http://yann.lecun.com/exdb/mnist/
--Featuers:
---60,000 training images.
---10,000 testing images.
---The images are divided into ten calsses from digit 1 to 9.
---Each Image is a grayscale image of 28x28 pixels.
***ImageNet [#afa1748e]
ImageNet is an image dataset used for training large scale object recognition models.
-https://www.image-net.org/challenges/LSVRC/index.php
-https://www.kaggle.com/c/imagenet-object-localization-challenge/data
--Featuers:
---Contains more than 14 million images according to the WordNet hierarchy.
---Images are annotated, identifying what objects pixelsresent in them.
---There are over 1 million images with bounding box annotations.
---The average image resolution on ImageNet is 468×387 pixels but is usually subsampled to 256×256 pixel to train deep learning models.
---There are more than 20,000 categories of objects that can be identified in the ImageNet dataset.
終了行:
[[MenuBar]]
CENTER:*DATA SET [#o624485f]
***CiFAR-10 [#m923cec7]
The CIFAR-10 is a labeled dataset of 60,000 images of animals and automobiles collected by Alex Krizhevsky, Vinod Nair, and Geoffery Ninton for training object classification models.
-https://www.cs.toronto.edu/~kriz/cifar.html
--Featuers:
---50,000 training images divided into 5 batches of 10,000 images each.
---10,000 testing images of a single batch.
---The images are divided into 10 classes; Airplane, Deer, Cat, Dog, Bird, Frog, Horse, ship, Truck.
---Each image is an RGB image of 32x32x3 pixels.
***CiFAR-100 [#h079723a]
The CIFAR-100 is a labeled dataset of 60,000 images of animals and automobiles collected by Alex Krizhevsky, Vinod Nair, and Geoffery Ninton for training object classification models.
-https://www.cs.toronto.edu/~kriz/cifar.html
--Featuers:
---50,000 training images divided into 5 batches of 10,000 images each.
---10,000 testing images of a single batch.
---The images are divided into 100 classes that are grouped into 20 super classes. Each image is labelled by its class and super class.
---Each image is an RGB image of 32x32x3 pixels.
***MNIST [#b61d656d]
MNIST is a dataset of handwritten digit images. It is commonly used for training and testing machine learning models.
-https://github.com/arXivTimes/arXivTimes/tree/master/datasets
-http://yann.lecun.com/exdb/mnist/
--Featuers:
---60,000 training images.
---10,000 testing images.
---The images are divided into ten calsses from digit 1 to 9.
---Each Image is a grayscale image of 28x28 pixels.
***ImageNet [#afa1748e]
ImageNet is an image dataset used for training large scale object recognition models.
-https://www.image-net.org/challenges/LSVRC/index.php
-https://www.kaggle.com/c/imagenet-object-localization-challenge/data
--Featuers:
---Contains more than 14 million images according to the WordNet hierarchy.
---Images are annotated, identifying what objects pixelsresent in them.
---There are over 1 million images with bounding box annotations.
---The average image resolution on ImageNet is 468×387 pixels but is usually subsampled to 256×256 pixel to train deep learning models.
---There are more than 20,000 categories of objects that can be identified in the ImageNet dataset.
ページ名: