Publishers accumulate a large image bank over time. However, reusing these images is difficult as they are not properly tagged and cannot be efficiently searched. AI provides the answer to more effective tagging of images and hence their reuse.
A picture, they say, is better than a thousand words. It is no wonder therefore that publishers accumulate a large body of images in their publishing systems over time. Publishers in fact spend a lot of time and mini fortunes acquiring these image banks. Reusing these images as is or repurposing them can thus improve both productivity (time) and finances (cost). However, in order to reuse, you need to be able to quickly search and find the right image with a minimum of effort. And in order for this search to work, the image need to be properly tagged. Image tags are descriptive words attached to images that can be used to search for and retrieve them. For example, an image could be tagged with the location it refers to, the names of the people in it. The colours in it, the size of the image, the camera or software that was used to generate it, the logos of companies if any, that are seen in it and so on. Theoretically, an image can have an unlimited number of tags describing various aspects. However, in practice it is a different story.
Writers and journalists are more focused on writing and publishing their content than on extensively and properly tagging their images. They tend to view tagging of images and other assets as an unnecessary evil and if forced, tend to give the bare minimum or even inane tags to get over the requirement. In fact, if they do not tag images or add inane tags, it may be better off than if they were to add wrong tags. Imagine what will happen of someone were to tag a photo of Theresa May, the British Prime Minister as Angela Markel, the German Chancellor. And later on, a writer writing about the British EU spat on Brexit adds this picture of Theresa May as that of Markel with a quote! To be fair to the writers, setting up and managing a good tagging system is the job of a qualified librarian than of a writer or an editor. Large publishers, particularly those with a history of print are likely to have well developed systems and even separate teams of people which can properly tag and catalogue these images. When it comes to web publishing and web content management systems, the story is different. Small teams and quickly changing audience and content environments mean that there may not even be the time and bandwidth to establish and update taxonomy dictionaries and image tags.
All this means that images in the image bank rarely get reused, and every time a new image is required, there is additional time and money spent creating one afresh rather than reusing or repurposing existing images.
With artificial intelligence-based image tagging, you can do away with the vagaries of manual tagging and implement a more consistent image tagging mechanism. AI systems of today can identify image mood, color, elements like water, sunrise etc. and features like mountains, towers, roads and so on. They can also identify images of popular figures and logos of popular companies. Some of the AI systems can even read text in images. All of these can now be stored as tags attached to images and can be used to search and locate images. For example, you could easily call up all images you have that have flowers in them, even if they are not named or tagged as such. Or you can search for all images having the Apple logo in them.
This means that writers now have to spend less time searching the internet for freely reusable images or giving briefings to digital artists or photographers. Publishers will also end up spend less money on images as more and more images get reused and repurposed.
Every image today comes with a lot of attached metadata. Digital photographs have data on the camera used, camera settings etc, (EXIF data). Digitally created or retouched images will have data like software used to create them or even the name of artists who worked on them.
Combining the tag data from AI systems with the machine readable meta data attached to images gives your web content management system an even more powerful image search functionality that writers and editors can easily use.
The Kreatio CMS uses Chitra AI, a central image store. When an image is uploaded to Chitra AI, it is automatically tagged for a variety of elements including personalities, colors and objects in the image as well as logs, text, mood and location. This rich set of tags is further enhanced with machine data available with the images. In addition to these, Chita AI can have human added tags as well as information like license types and artist information attached to the image.
Every time an image is added to an article in the Kreatio CMS, Chitra AI tracks which website it is being published to and ads it to an album of the same name. Further, it tracks the URLs of the articles to which it has been attached as well as the number of times it has been uses.
Thus, publishers can share images across their publications with the confidence that writers can easily reuse them while being aware of where all they have been used previously.
Those who work for long hours on an application would love to customize it to their preferences and comfort. The Kreatio CMS lets authors and other users customize the order of its modules and workspaces to their liking Every one of us have our own comfort zones and ways of working. Therefore, it is only natural that we attempt to organize our work areas to match our preferences and tastes. For an author or editor in a digital media house, the Content Management System (CMS) is one piece of software that they spent a lot of time with. Given today's complex business environment, CMSes tend to...