Skip to main content

Image augmentation auxiliary tool

Project description

AugWrap

PyPI version

Installation

$ pip install augwrap

Basic Usage

  1. Base dataset을 선택합니다. Pytorch 기반 데이터셋을 만들려면 TorchBaseDataset을 사용합니다. Tensorflow 기반 데이터셋을 만들려면 TFBaseDataset을 사용합니다.

    import augwrap as aw
    
    dataset = aw.data.TorchBaseDataset(images, labels, classes)
    
  2. augwrap.data 안의 모듈을 이용하여 학습 전 데이터를 처리합니다.

    dataset = aw.data.LoadImages(dataset)
    dataset = aw.data.ResizeImages(dataset, (256, 256))
    dataset = aw.data.OneHotLabels(dataset)
    
  3. 데이터 로더를 생성하여 학습에 사용합니다.

    from torch.utils.data import DataLoader
    
    data_loader = DataLoader(
        dataset,
        batch_size = 16,
        shuffle = False,
        num_workers = 4
    )
    

Augmentations

잘 알려진 어그멘테이션 도구인 Albumentations을 활용할 수 있도록 만들었습니다.

augwrap.data.Augmentations의 생성자 인자로 base dataset에서 파생된 객체와 Albumentations 객체를 받습니다.

Albumentations와 함께 사용할 수 있는 어그멘테이션 모듈을 augwrap.augmentations에 추가했습니다.

import albumentations as A
import augwrap as aw

augmentations = A.Compose([
        A.RandomRotate90(p=1),
        A.GridDistortion(p=0.8),
        A.GaussNoise(p=0.75),
        aw.augmentations.CutMix(dataset, p=0.8),
])
dataset = aw.data.Augmentations(dataset, augmentations)

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

augwrap-0.0.1a2.tar.gz (11.2 kB view details)

Uploaded Source

Built Distribution

augwrap-0.0.1a2-py3-none-any.whl (9.4 kB view details)

Uploaded Python 3

File details

Details for the file augwrap-0.0.1a2.tar.gz.

File metadata

  • Download URL: augwrap-0.0.1a2.tar.gz
  • Upload date:
  • Size: 11.2 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.6.0 importlib_metadata/4.8.2 pkginfo/1.8.1 requests/2.26.0 requests-toolbelt/0.9.1 tqdm/4.62.3 CPython/3.9.7

File hashes

Hashes for augwrap-0.0.1a2.tar.gz
Algorithm Hash digest
SHA256 09481c4aed6dbe17a553b686c80f44ff5e7e05d5a8b305f076c615fc745516b4
MD5 5902fafff20470770d302ab3032ba34c
BLAKE2b-256 cc8b8e374ea8b6935bc521458cda1d90b5b7e3e418b0fc10f0e51fc78a43500d

See more details on using hashes here.

File details

Details for the file augwrap-0.0.1a2-py3-none-any.whl.

File metadata

  • Download URL: augwrap-0.0.1a2-py3-none-any.whl
  • Upload date:
  • Size: 9.4 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.6.0 importlib_metadata/4.8.2 pkginfo/1.8.1 requests/2.26.0 requests-toolbelt/0.9.1 tqdm/4.62.3 CPython/3.9.7

File hashes

Hashes for augwrap-0.0.1a2-py3-none-any.whl
Algorithm Hash digest
SHA256 1b28f7166e13432a31ab3bbc0622138d03d88b3351e70d7a5c693c860d7fc719
MD5 f904e64976f0196c3d238b2e5c01b46b
BLAKE2b-256 55383440093b1e62d997c18a5e364287f456200e96e9c2738aaebe908477eed2

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page