Asset Bank's new Auto Tagging feature extracts keywords from images at point of upload or update, saving time and effort when you need to tag lots of assets at once.
Auto Tagging uses Amazon Rekognition, an intelligent toolkit which identifies objects, faces and reads text from your images and videos. Amazon Rekognition is based on the same proven, highly scalable, deep learning technology developed by Amazon’s computer vision scientists to analyse billions of images and videos daily.
Types of Auto Tagging
Auto Tagging can be configured to use up to three different methods of keyword extraction:
1. Object - extracts words based on objects found within your images & videos.
2. OCR (Optical Character Recognition) - extracts phrases found in images or videos.
3. Face - if a human face is found in an image or video asset, this type will add the word 'Face'.
Enabling & Configuring of Auto Tagging
Using Auto Tagging
Once enabled & configured, Auto Tagging can be used during a single or bulk upload. It can also be used to tag assets that already exist on your Asset Bank by performing an embedded metadata re-extraction in Admin > Attributes > Embedded Data Mappings > Rerun metadata extraction for existing assets