Over the last few months, we have been making significant changes to the ImageKit Media Library to allow you to use it better as a digital asset management platform. Advanced search queries, embedded image metadata parsing, complete folder uploads, built-in image editor, and consistent developer APIs have been significant steps in that direction.
During this product evolution, we received feedback from the users about certain third-party services they were using or tasks they were doing during their image upload or editing process.
There was a scope to significantly automate these tasks with the help of AI and existing industry-leading services.
Building on this learning, the latest addition to the Media Library trumps every other past feature in opening up incremental use cases to be solved within the ImageKit platform.
Introducing ImageKit Media Library Extensions - developer-friendly, third-party integrations to enable powerful yet simple workflows within the ImageKit digital asset management system. The aim of these extensions, the ones that we have launched or get added in the future, is to eliminate the need for you to use multiple services for any media-related task and have it all work seamlessly within ImageKit.
To begin with, we have added integrations with three very useful services.
Let's take a look at them, along with the use cases that they solve.
1. Extensions for Google Cloud Vision and AWS Rekognition
Automatic image tagging to enable better digital asset management and search
With the enormous amount of media content created in-house or by our users, it is impossible to search for content just by the file name or the folder name.
We need ways to help us make sense of the actual "content" of an asset than just its name.
With the developments in AI, understanding an image's content is now easier than ever. Two industry-leading APIs that help label commonly known objects in an image are Google Cloud Vision and AWS Rekognition.
For example, Google Cloud Vision identifies the objects 'Bottle', 'Cola', 'Soft Drink', 'Plants', and more with over 70% certainty in the image below.
Now with ImageKit, at the time of image upload or edit via the API, you can now optionally trigger the Cloud Vision or Rekognition APIs to generate tags for the content in the image.
These tags will then get stored with the image along with any manual tags that you might have added.
You can find the API documentation for using these auto-tagging extensions here along with the possible extension options.
These extension integrations tag the image asynchronously, i.e., your image uploads are still fast, and the tags get added "eventually" and not in real-time.
Upcoming: With your image content now tagged, you will be able to search for images based on their actual content. You would be able to query for images containing specific AI tags, and ImageKit will pull out a list of images matching your criteria. This update is likely to be released by the 1st week of November 2021. Watch out this space for updates.
2. Extension for Remove.bg
AI-powered background removal and replacement for better visual experiences
Delivering a consistent visual experience on the entire website is the holy grail for every business online.
And when you are dealing with images coming in from multiple sources or for different products, especially if they are coming directly from your users, it gets difficult to maintain that visual consistency.
Graphic teams have to edit the images manually, remove their existing backgrounds and replace them with what they use for their brand.
ImageKit now provides integration with the best AI-powered background removal service - Remove.bg. You can now upload or update an image in the Media Library and have its background removed or replaced automatically with an exceptionally high degree of accuracy.
You can read the API documentation for the background removal extension here along with the possible extension options.
The image after background removal gets uploaded to your media library, and you can perform transformations and optimizations on it like any other image.
For example, we used this extension to remove the background from the Coca-Cola bottle image and then gave it a solid background using ImageKit's parameters.
## Original image https://ik.imagekit.io/ikmedia/bgremove/original_bottle.jpeg ## Background Removed https://ik.imagekit.io/ikmedia/bgremove/bgremoved_bottle.jpg?tr=w-600,bg-DDDDDD
You can use this extension to homogenize the backgrounds for your product images or user-uploaded images.
How can you try these extensions? And how will the usage get billed?
Since extensions perform different tasks with varying difficulty and utility, their consumption needs to be billed differently.
But, to reduce the complexity, the consumption will be measured in terms of "Extension Units". Each use of an extension would consume a certain number of units, as mentioned here, which would be in line with the task performed by the extension.
For you to start using these extensions, each plan now has inclusions for Extension units. You will be able to find them on the plans page in your dashboard and on the website.
For paid users, units consumed beyond the plan inclusions will be charged on a usage basis as described in their plan.
Starting 1st November 2021, usage of extensions will be charged across all plans, with the exception of our Forever Free plan. For any queries regarding this, please reach out to our support team.
What's the way forward?
These three extensions represent just the beginning of a vast set of functionalities to be made available inside the Media Library. Based on customer feedback, we will continue to integrate more third-party services with ImageKit.
We are also working on getting these third-party services available not just via the APIs but also inside the ImageKit dashboard.
Want to know more?
If you have a service you would like us to integrate or want a custom plan for your requirement or face any challenges in using the extensions, please write to us at firstname.lastname@example.org or start a live chat from your dashboard. We would really love to hear your feedback.