Skip to main content

An open source toolkit for evaluating the natural robustness of computer vision algorithms.

Project description

Natural Robustness Toolkit (NRTK)

The nrtk package is an open source toolkit for evaluating the natural robustness of computer vision algorithms to various perturbations, including sensor-specific changes to camera focal length, aperture diameter, etc. Functionality is provided through Strategy and Adapter patterns to allow for modular integration into systems and applications.

We have also created the nrtk-jatic package to support AI T&E use-cases and workflows, through interoperability with the maite library and integration with other JATIC tools. Users seeking to use NRTK to perturb MAITE-wrapped datasets or evaluate MAITE-wrapped models should start with the nrtk-jatic package.

Installation

The following steps assume the source tree has been acquired locally.

Install the current version via pip:

pip install nrtk

Alternatively, you can also use Poetry:

poetry install --sync --with dev-linting,dev-testing,dev-docs

See here for more installation documentation.

Getting Started

We provide a number of examples based on Jupyter notebooks in the ./examples/ directory to show usage of the nrtk package in a number of different contexts.

Contributions are welcome! See the CONTRIBUTING.md file for details.

Documentation

Documentation snapshots for releases as well as the latest master are hosted on ReadTheDocs.

The sphinx-based documentation may also be built locally for the most up-to-date reference:

# Install dependencies
poetry install --sync --with dev-linting,dev-testing,dev-docs
# Navigate to the documentation root.
cd docs
# Build the docs.
poetry run make html
# Open in your favorite browser!
firefox _build/html/index.html

Developer tools

pre-commit hooks pre-commit hooks are used to ensure that any code meets all linting and formatting guidelines required. After installing, this will always run before committing to ensure that any commits are following the standards, but you can also manually run the check without committing. If you want to commit despite there being errors, you can add --no-verify to your commit command.

Installing pre-commit hooks:

# Ensure that all dependencies are installed
poetry install --sync --with dev-linting,dev-testing,dev-docs
# Initialize pre-commit for the repository
poetry run pre-commit install
# Run pre-commit check on all files
poetry run pre-commit run --all-files

Contributing

License

Apache 2.0

Contacts

Principal Investigator: Brian Hu (Kitware) @brian.hu

Product Owner: Austin Whitesell (MITRE) @awhitesell

Scrum Master / Tech Lead: Brandon RichardWebster (Kitware) @b.richardwebster

Deputy Tech Lead: Emily Veenhuis (Kitware) @emily.veenhuis

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

nrtk-0.15.1.tar.gz (34.8 kB view details)

Uploaded Source

Built Distribution

nrtk-0.15.1-py3-none-any.whl (56.6 kB view details)

Uploaded Python 3

File details

Details for the file nrtk-0.15.1.tar.gz.

File metadata

  • Download URL: nrtk-0.15.1.tar.gz
  • Upload date:
  • Size: 34.8 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: poetry/1.8.4 CPython/3.9.20 Linux/6.5.0-1021-aws

File hashes

Hashes for nrtk-0.15.1.tar.gz
Algorithm Hash digest
SHA256 3d764edd9ee7bb368695ec445dd198a0f25dfbe32445f0e9db0adb6322d00bbb
MD5 e2d0f56eb2248f3a0b4c88713305040c
BLAKE2b-256 1cb6391bfd2d8f2ab18481aadce263cc242198b187e9b7f653a139b8a4daa348

See more details on using hashes here.

File details

Details for the file nrtk-0.15.1-py3-none-any.whl.

File metadata

  • Download URL: nrtk-0.15.1-py3-none-any.whl
  • Upload date:
  • Size: 56.6 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: poetry/1.8.4 CPython/3.9.20 Linux/6.5.0-1021-aws

File hashes

Hashes for nrtk-0.15.1-py3-none-any.whl
Algorithm Hash digest
SHA256 255827ee948ff3ee35a3c620acdf4709ae88192b3e9b8efc8dea4dede330fb65
MD5 186d38c84d43f53819e0502ee734ae47
BLAKE2b-256 cbdb698885aee97ff2618dc5a917decb0fdcedf657355043bfd7317f753d8060

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