From version 6 onward, AEM uses machine learning algorithms. These are the major AEM features built on machine learning.
| Feature | What it does |
|---|---|
| Smart Tags | Deep-learning-based image tagging |
| Enhanced Smart Tags | Lets you train smart tagging to recognize your own concepts, such as logos |
| Smart Crop | Identifies the most important parts of an image and crops accordingly |
| Smart Swatch | Identifies the most representative region of an image for a color swatch |
| Smart Translation Search | Translates search queries on the fly to support search in multiple languages |
| Smart Summarization | Summarizes a text by identifying its most important sentences |
| Expert Scoring | Identifies experts among community users |
| Spam detection for communities | Flags potential spam comments automatically |
| Smart Forms Conversion (tech preview in 6.4) | Semi-automates converting forms from any format to AEM forms using deep learning |
| Smart Layout (tech preview in 6.4) | Suggests layouts for different audiences based on KPIs and an understanding of the content |
AEM 6.5 extended these capabilities to video with Smart Crop, best-frame selection and Smart Tags for video — see AEM 6.5 new features in Assets.