Skip to content

Latest commit

 

History

History
54 lines (37 loc) · 1.69 KB

README.md

File metadata and controls

54 lines (37 loc) · 1.69 KB
Raw

Computer Vision Utilities

This is an attempt to put together a collection of resources and scripts I put together while working on various Computer Vision projects, hope they can help you save your time if you come across tasks these resources can handle.

Setup

In Terminal

git clone https://github.com/r3stl355/cv-utils.git
cd cv-utils

I often use Conda environments for my projects but any other equivalent will do. Here is an example of setting up a Conda environment to test the scripts

  • First, install Conda (my personal preferece is Miniconda)
  • Then
conda create -n cv-utils python=3.7
conda activate cv-utils
pip install -r requirements.txt

Running Jupyter Notebooks

Most of the resources here are implemented as Jupyter Notebooks, with others being just loose script files.

Run a Jupyter server and check the Notebook(s) of interest in the Jupyter instance that opens in your browser (if it does not launch automatically, just copy/paste the URL shown in the Terminal to your browser address bar)

jupyter notebook

Convert a dataset in Pascal VOC format to RecordIO format

Check out voc_2_rec.ipynb Notebook for details.

Convert a dataset in MS COCO format to RecordIO format

Check out coco_2_rec.ipynb Notebook for details.

Shuffle and split an .lst file into train and val files

Script is using fixed split of 90/10, adjust as needed to fit your needs within the script or extend the script to accept more parameters. Third parameter determines the shuffle tool to use, e.g. gshuf on Mac OS installed as part of coreutils

source split_train_val.sh data/voc_like_sample test.lst gshuf

and so on...