distort --me

CLI utility for augmentation and preprocessing images.

---

Table of Contents

• Installation
• Installation from source
• Features and commands
• Usage
• License

---

Installation

$ pip install distortme

Installation from source

Install poetry to build package

$ pip install poetry

Clone

Clone repo to your local machine

$ git clone https://github.com/rahowa/distortme.git

Setup

Now install package to local enviroment with poetry

$ cd distortme
$ poetry install

Check installation

$ distortme --help

---

Features and commands

Commands:

• distortme info Get info about certain dataset
• distortme unpack Unpack any archive with one command
• distortme show Show image inside terminal for fast checking
• distortme augs Perform augmentations for images before training
• distortme voc2coco Convert PASCAL VOC dataset format to COCO format
• distortme coco2voc Convert COCO dataset format to PASCAL VOC format
• distortme split Sort by folders images with certain descriptor in name
• distortme map Apply custom dataset preprosesing with multiple prosesses
• distortme torle distortme fromrle Convert segmentation masks to RLE-encoding and back
• distortme tohd5 distortme fromhd5 Compress dataset to HDF5 format and decompress back
• distortme download Allows to download most popular datasets to given folder without any code
• distortme label Create labels for Face detection, Object detection [IN PROGRESS] and Classification [IN PROGRESS] tasks for your custom dataset

Options:

• --install-completion: Install completion for the current shell.
• --show-completion: Show completion for the current shell, to copy it or customize the installation.
• --help: Show this message and exit.

Documentation

For more information check DOCS.md file

Usage

Use distortme as a regular CLI app:

$ distortme [OPTIONS] COMMAND [ARGS]...

---

Created with

License

MIT license