Responsive design patterns are state-of-the-art for websites, landing pages, and online product catalogs, requiring images and videos to be placed once and resized or cropped based on devices and other criteria. Maximizing the impact of each visual is crucial to attracting users' attention and sales conversion.

Focal point cropping, with each image cropped to keep the focus on the key elements at each size, is now easier than ever. You select the AI-delivered main point of interest (focal point) and crop automatically around it, focusing on the selected key element.

AI tagging in Picturepark will therefore propose additional focal points, which you can use in the media editor for cropping. These AI-delivered focal points will consider all possible areas of interest for you - without the need to set them yourself.

You find various areas e.g. House, Tree, or Plant from the image. Depending on the model you choose AI can also propose brand names, text, or demographics. AI technology will add such focal points to a layer “Image Analytics”, from which you can pick the ones to apply as a cropped onto your images in the Media Editor.

You can further define the desired height and width - the output dimensions (e.g. 16:9) and other settings. Given the requested width and height, the resulting image preserves as much of the region of interest as possible.

Further Changes

• You can use a new crop parameter "auto" in the embed URL to set the crop rectangle to the ideal dimensions for the selected focal point.
  

Limitation

• AI Tagging in Picturepark can either return focal points (regions with names) or concepts (keywords/tags). It is not possible to configure both tagging options for the same instance, it is either focal points or tags.

Business Rules in Picturepark automate your data management and are now extended with user-centered conditions. These new user-centered conditions such as the “User in User Roles” condition allow you to automate processes based on actions performed by particular users or user roles, such as::

• Assign default permission when your agency uploads the content.

• Assign default tags and keywords when your photographers upload.

• Send out notifications to specific users or user roles when your external suppliers change metadata.

• Check if product content was made public by a product manager and otherwise revoke permission

• Check if sales content was validated by the product manager and request validation.

• And many more.

Further Changes in Business Rules

• A new variable NOW is available for automation to check whether a specific release date or embargo date is approaching. For instance, you can calculate “NOW + 5 Days” to check for an embargo date to expire five days from now.

• A new “contains” condition checks if any input contains specific text, e.g. if the filename contains product, to automatically apply the product layer. Previously, this process was only possible with complex RegEx before and has now become simplified.

• A name and description field was added to describe each step of a business rule, which helps you better document a business rule directly where you configure it, hence supporting your corporate business governance strategy.

• The condition to check if a layer was changed now also supports checking if multiple layers of a Content Item had changed.

With this release, the Display Pattern Editor offers an exact preview of what your Display Patterns look like - without rendering all content items in Picturepark which did consume some time while you waited.

This is especially helpful for Virtual Items and Thumbnail Previews (overlays) where you can show graphical information in the top left, bottom left, and bottom right corners, such as icons, text in different colors, and more.

We switched our deployment to Kubernetes which reduces internal efforts for managing the architecture, unavailability during releases and maintenance windows, and further improves the scalability of our Picturepark Cloud.

Kubernetes is a compact, portable open-source platform for managing applications like Picturepark, originally developed by Google.

Benefits of Kubernetes which we have started to utilize include:

• Scalability: Scale-up certain services based on load indicators, and more.

• Load Balancing: Balance incoming traffic load to all containers and systems evenly.

• Automatic Packing: Handling any number of containers on its own.

• Self Healing: Replace a broken container in due time to maintain the capacity and bandwidth of the system.

• Powerful CI/CD (which means continuous integration and development): Kubernetes will gradually deploy to the entire swarm of containers.

• Storage orchestration: Scale storage systems of all kinds - including local storage, public cloud providers, etc.