[Commons-l] Fwd: [Wikitech-l] Using computer vision to categorize images at Commons