Technology sharing

한어Русский языкEnglishFrançaisIndonesianSanskrit日本語DeutschPortuguêsΕλληνικάespañolItalianoSuomalainenLatina

1. Data amplificationem

1. Exsistentia notitias addere ad pluris diversitatem pertinentes, ut addens sonum diversum colorum et colorem et figuram picturae mutans.

2. Consectetur notitia generatur online

3. genus amplificationem:

(1) Flip

(2) Secare

(3) Color

(IV) exacue et addere maculae

2. Libri

1. Augmentatio data diversitatem obtinet per informationes informandos, per hoc quod exemplar magis generale facit.

2. Communia imaginis amplificationes includunt flippingem, incisionem, ac decolorationem

3 Just mutare notitia paro

4. Ut accurate eventa in processu praedictionis obtineamus, solemus tantum imaginum augmentationem in exemplorum institutione praestare, et augmentationem imaginis non uti per processum praedictum cum incertis operationibus.

5. Alta doctrinarum compage multas varias imaginis augmentationes methodos praebet, quae simul applicari possunt.

3. Code

1. Augmentum modusaug

def apply(img, aug, num_rows=2, num_cols=4, scale=1.5):
    Y = [aug(img) for _ in range(num_rows * num_cols)]
    d2l.show_images(Y, num_rows, num_cols, scale=scale)

2. Flip

apply(img, torchvision.transforms.RandomHorizontalFlip())
apply(img, torchvision.transforms.RandomVerticalFlip())

3. secans

#面积为原始面积10%到100%的区域，宽高比从0.5～2之间随机取值
shape_aug = torchvision.transforms.RandomResizedCrop(
    (200, 200), scale=(0.1, 1), ratio=(0.5, 2))
apply(img, shape_aug)

4. splendor

apply(img, torchvision.transforms.ColorJitter(
    brightness=0.5, contrast=0, saturation=0, hue=0))

5. Color tonus

apply(img, torchvision.transforms.ColorJitter(
    brightness=0, contrast=0, saturation=0, hue=0.5))

claritas (brightness) Contra (contrast), satietatem (saturation) Et colore (hue）

6. Mix

augs = torchvision.transforms.Compose([
    torchvision.transforms.RandomHorizontalFlip(), color_aug, shape_aug])
apply(img, augs)