Metadata-Version: 2.1
Name: quevedo
Version: 1.1.0
Summary: Tool for managing datasets of images with compositional semantics
Home-page: https://github.com/agarsev/quevedo
License: OSL-3.0
Keywords: machine learning,computer vision,sign language,corpus linguistics,dataset
Author: Antonio F. G. Sevilla
Author-email: afgs@ucm.es
Requires-Python: >=3.7,<4.0
Classifier: License :: OSI Approved
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.7
Classifier: Programming Language :: Python :: 3.8
Classifier: Programming Language :: Python :: 3.9
Provides-Extra: force_layout
Provides-Extra: web
Requires-Dist: click (>=8,<9)
Requires-Dist: flask (>=1.1.2,<2.0.0); extra == "web"
Requires-Dist: forcelayout (>=1.0.6,<2.0.0); extra == "force_layout"
Requires-Dist: pillow (>=8,<9)
Requires-Dist: toml (>=0.10.2,<0.11.0)
Project-URL: Documentation, https://agarsev.github.io/quevedo
Project-URL: Repository, https://github.com/agarsev/quevedo
Description-Content-Type: text/markdown

![Quevedo Logo](quevedo/logo.png)

# Quevedo

Quevedo is a python tool for managing datasets of images with compositional
semantics, with a focus on the training and evaluation of machine learning
algorithms on these images.

Quevedo is part of the [VisSE project](https://www.ucm.es/visse). The code can
be found at [GitHub](https://github.com/agarsev/quevedo), and [detailed
documentation here](https://agarsev.github.io/quevedo).

## Features

- Dataset management, including hierarchical dataset organization, subset
    partitioning, and semantically guided data augmentation.
- Structural annotation of source images using a web interface, with support for
    different users and the live visualization of data processing scripts.
- Deep learning network management, training, configuration and evaluation,
    using [darknet].

## Installation

Quevedo requires `python >= 3.7`, and can be installed from
[PyPI](https://pypi.org/project/quevedo/):

```shell
$ pip install quevedo
```

Or, if you want any extras, like the web interface:

```shell
$ pip install quevedo[web]
```

Or directly from the wheel in the [release
file](https://github.com/agarsev/quevedo/releases):

```shell
$ pip install quevedo-1.1.0-py3-none-any.whl[web]
```

To use the neural network module, you will also need [to install
darknet](https://agarsev.github.io/quevedo/latest/nets/#installation).

### Development

To develop on quevedo, we use [poetry] as our environment, dependency and build
management tool. In the quevedo code directory, run:

```shell
$ poetry install
```

Then you can run quevedo with

```shell
$ poetry run quevedo
```

## Usage

To create a dataset:

```shell
$ quevedo -D path/to/new/dataset create
```

Then you can **cd** into the dataset directory so that the `-D` option is not
needed.

To see information about a downloaded dataset:

```shell
$ quevedo info
```

To launch the web interface (you must have installed the "web" extra):

```shell
$ quevedo web
```

For more information, and the list of commands, run `quevedo --help` or `quevedo
<command> --help` or see [here](https://agarsev.github.io/quevedo/latest/cli/).

## Dependencies

Quevedo makes use of the following open source projects:

- [python 3]
- [poetry]
- [darknet]
- [click]
- [flask]
- [preactjs]

Additionally, we use the [toml] and [forcelayout] libraries, and build our
documentation with [mkdocs].

## About

Quevedo is licensed under the [Open Software License version
3.0](https://opensource.org/licenses/OSL-3.0).

The web interface includes a copy of [preactjs] for ease of offline use, distributed
under the [MIT License](https://github.com/preactjs/preact/blob/master/LICENSE).

Quevedo is part of the project "Visualizando la SignoEscritura" (Proyecto VisSE,
Facultad de Informática, Universidad Complutense de Madrid) as part of the
program for funding of research projects on Accesible Technologies financed by
INDRA and Fundación Universia.

### VisSE team

- [Antonio F. G. Sevilla](https://github.com/agarsev) <afgs@ucm.es>
- [Alberto Díaz Esteban](https://www.ucm.es/directorio?id=20069)
- [Jose María Lahoz-Bengoechea](https://ucm.es/lengespyteoliter/cv-lahoz-bengoechea-jose-maria)

[darknet]: https://pjreddie.com/darknet/install/
[poetry]: https://python-poetry.org/
[python 3]: https://www.python.org/
[click]: https://click.palletsprojects.com/
[flask]: https://flask.palletsprojects.com/en/2.0.x/
[preactjs]: https://preactjs.com/
[toml]: https://pypi.org/project/toml/
[forcelayout]: https://pypi.org/project/forcelayout/
[mkdocs]: https://www.mkdocs.org/

