Will probably shift to something like Amazon Rekognition / Google Cloud Vision next...

Will probably shift to something like Amazon Rekognition / Google Cloud Vision next weekend, to get deeper context about the image