Health-related spam images

Published: 8 February 2021| Version 2 | DOI: 10.17632/vm5msmw257.2
Contributor:
NIDDAL IMAM

Description

300 labeled images collected from Twiiter Arabic hashtags and used for fine-tuning a text loclization (pixel_link) part of OCR system. The datasets were used in our published paper "Detecting Spam Images with Embedded Arabic Text in Twitter".

Files

Steps to reproduce

The images were collected from Twitter and labeled using Labelimg (https://github.com/tzutalin/labelImg). Please refer to our repo (https://github.com/niddal-imam/End-2-End-image-spam-detector-pixel_link) for more details.

Institutions

University of York

Categories

Optical Character Recognition, Twitter

Licence