Large scale image dataset visiualization tool in python

Show Data: Show your dataset in web browser!

Show Data is to generate html tables for large scale image dataset, especially for the dataset in remote server. It provides some useful commond line tools and fully customizeble API reference to generate html table different tasks.

Examples

It supports filtering data by text, sorting data by coloum values and pagination.

VeRiWild dataset

ReID Strong baseline Results

Install

pip install showdata

Command Line Tools

Basic usage

Open a file server (replace the python -m http.server)

showdata server -p <port, default 8000> -h <host, default 0.0.0.0>

Visualize images in a folder:

showdata show -i <folder_path> -o <html_path>

Visualize images from pickle or json file:

showdata show -i example.pkl -o <html_path>

Compare images with the same name from different folders

showdata compare <path-a> <path-b> <path-c> -o <output-path>

All string values ends with png/jpg/jpeg will be rendered to images. Others are rendered to text.

Image Dict

If you pass a json file, you can set the image values to a dict to provide more customizeble configurations.

For example,

[
    {
        "name": "001",
        "img": {
            "src": "images/0001.jpg",
            "text": "The text will be shown on the top of the image",
            "style": "border: 2mm solid green"
        }
    }
    {
        "name": "002",
        "img": {
            "src": "images/0002.jpg",
            "text": "blabla",
            "style": "border: 2mm solid red"
        }
    }
]

Run python -m http.server and open your web browser to check the generated web page.

API

from showdata import generate_html_table
data = [
    {
        "idx": 1,
        "label": 'cat',
        "img": 'images/cat.jpg',
        "mask": 'images/cat_mask.png',
    },
    {
        "idx": 2,
        "label": 'dog',
        "img": 'images/dog.jpg',
        "mask": 'images/dog_mask.png',
    },
]
generate_html_table(data, output_path='index.html')