Data

Tools

Indexes

Applications

Experiments

Open Source Science

icgen

Image Classification Dataset Generator

https://github.com/automl/icgen

Science Score: 36.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • ○
    CITATION.cff file
  • ✓
    codemeta.json file
    Found codemeta.json file
  • ✓
    .zenodo.json file
    Found .zenodo.json file
  • ○
    DOI references
  • ○
    Academic publication links
  • ✓
    Committers with academic emails
    1 of 2 committers (50.0%) from academic institutions
  • ○
    Institutional organization owner
  • ○
    JOSS paper metadata
  • ○
    Scientific vocabulary similarity
    Low similarity (9.3%) to scientific vocabulary
Last synced: 10 months ago · JSON representation

Repository

Image Classification Dataset Generator

Basic Info
  • Host: GitHub
  • Owner: automl
  • License: mit
  • Language: Python
  • Default Branch: master
  • Homepage:
  • Size: 101 KB
Statistics
  • Stars: 3
  • Watchers: 7
  • Forks: 3
  • Open Issues: 0
  • Releases: 0
Created about 6 years ago · Last pushed over 4 years ago
Metadata Files
Readme Contributing License Citation

README.md

ICGen

Installation

pip install icgen

for a development installation see CONTRIBUTING.md

Usage

Sampling Datasets

python import icgen dataset_generator = icgen.ICDatasetGenerator( data_path="datasets", min_resolution=16, max_resolution=512, max_log_res_deviation=1, # Sample only 1 log resolution from the native one min_classes=2, max_classes=100, min_examples_per_class=20, max_examples_per_class=100_000, ) dev_data, test_data, dataset_info = dataset_generator.get_dataset( dataset="cifar10", augment=True, download=True )

The augment parameter controls whether the original dataset is modified.

Options only affect sampling with augment=True and the min max ranges do not filter datasets.

The data is left at the original resolution, so it can be resized under user control. This is necessary to for example avoid resizing twice which can hurt performance.

You can also sample from a list of datasets python dataset_generator.get_dataset(datasets=["cifar100", "emnist/balanced"], download=True)

We provide some lists of available datasets python import icgen icgen.DATASETS_TRAIN icgen.DATASETS_VAL icgen.DATASETS_TEST icgen.DATASETS or on the commandline you can get the names with python -m icgen.dataset_names

Downloading Datasets Before Execution

To download datasets ahead of time you can run

python -m icgen.download --data_path DATA_PATH --datasets D1 D2 D3

or directly download a complete group

python -m icgen.download --data_path DATA_PATH --dataset_group GROUP # all, train, dev, test

Alternatively, you can also use the download=True flag of the dataset_generator.get_dataset function.

Reconstructing and Distributing Tasks

In distributed applications it may be necessary to sample datasets on one machine and then use them on another one. Conversely, for reproducibility it may be necessary to store the exact dataset which was used. For these cases icgen uses a dataset identifier which uniquely identifies datasets.

License

MIT

Owner

  • Name: AutoML-Freiburg-Hannover
  • Login: automl
  • Kind: organization
  • Location: Freiburg and Hannover, Germany

GitHub Events

Total
  • Member event: 1
Last Year
  • Member event: 1

Committers

Last synced: about 1 year ago

All Time
  • Total Commits: 14
  • Total Committers: 2
  • Avg Commits per committer: 7.0
  • Development Distribution Score (DDS): 0.071
Past Year
  • Commits: 0
  • Committers: 0
  • Avg Commits per committer: 0.0
  • Development Distribution Score (DDS): 0.0
Top Committers
Name Email Commits
Danny s****d@c****e 13
Fábio Ferreira f****a@m****g 1
Committer Domains (Top 20 + Academic)
mailbox.org: 1 cs.uni-freiburg.de: 1

Issues and Pull Requests

Last synced: 11 months ago

All Time
  • Total issues: 0
  • Total pull requests: 0
  • Average time to close issues: N/A
  • Average time to close pull requests: N/A
  • Total issue authors: 0
  • Total pull request authors: 0
  • Average comments per issue: 0
  • Average comments per pull request: 0
  • Merged pull requests: 0
  • Bot issues: 0
  • Bot pull requests: 0
Past Year
  • Issues: 0
  • Pull requests: 0
  • Average time to close issues: N/A
  • Average time to close pull requests: N/A
  • Issue authors: 0
  • Pull request authors: 0
  • Average comments per issue: 0
  • Average comments per pull request: 0
  • Merged pull requests: 0
  • Bot issues: 0
  • Bot pull requests: 0
View more stats
Top Authors
Issue Authors
Pull Request Authors
Top Labels
Issue Labels
Pull Request Labels

Packages

  • Total packages: 1
  • Total downloads:
    • pypi 29 last-month
  • Total dependent packages: 0
  • Total dependent repositories: 1
  • Total versions: 2
  • Total maintainers: 1
pypi.org: icgen

Image Classification Dataset Generator

  • Versions: 2
  • Dependent Packages: 0
  • Dependent Repositories: 1
  • Downloads: 29 Last month
Rankings
Dependent packages count: 10.1%
Forks count: 19.2%
Dependent repos count: 21.6%
Stargazers count: 25.1%
Average: 28.2%
Downloads: 65.0%
Maintainers (1)
DaStoll
Last synced: 11 months ago

Dependencies

poetry.lock pypi
  • 118 dependencies