AEM Visual & Semantic Search

We are pleased to announce the release of AEM Visual & Semantic Search.

Starting from real customer needs, CodeLand Research & Development team has closed the implementation of a new AI feature in the field of search images by similarity and natural language description.
We use CLIP (Contrastive Language Image Pretraining) developed by Open AI to generate embeddings for images and text queries.
No need to send out images outside of customer’s organization and everything perfectly integrated in AEM with the possibility to take advantage of 2 pre-built components:

• Search textfield with capability to upload an image to search
• Image gallery with possibility to discover similar image

The solution can be easily integrated in any AEM application at scale.

Possible case studies:

• Supporting B2C and B2B Commerce websites
• Seach of products in commerce sites starting from a real life photo taken by mobile/tablet
• Product identification from ambient images
• Search of similar / alternative images starting from a given one
• Search of images based on textual description