A re-implementation of the genre-scraper Robbie Barrat included in art-DCGAN. Wikiart has since updated their website so that images are dynamically loaded. This repo uses selenium to gather links to individual artworks and requests to download the images.

pip install -r requirements.txt

Set pages on line 27 to the number of pages you want to scrape. There are 60 images per page, and the max number of pages is also 60.

Change genre-to-scrape on line 27 to any one of the genre names commented below.

Run using python3 genre-scraper.py.

• Progress bar for monitoring downloader's progress
• Command-line arguments for setting genre, style and output dir