Xcavator is a new type of search engine which utilizes visual clues that you provide to identify and extract similar pictures from large groups of digital images.
Similar to another existing Flickr visual search tool called Retrievr, which allows you to make a visual sketch and then find digital images based on that reference, this new visual search engine called Xcavator searches the huge Flickr images database by utilizing as search criteria specific micro areas within your selected picture.
To provide your visual search clues you first select a reference image, and then identify within it visual elements that you want to become your visual search query specifications. You do this by clicking one or more individual points on the reference image while excavator samples them and their surrounding are to determine the unique traits, characteristics and visual patterns present in each of the reference points. Then in real-time Xcavator displays as many "matching" images to your visual specifications as possible.
To note is the fact that you can progressively add new specs and eliminate them while you see corresponding visual matches being displayed on-screen in real-time. This allows the user to quickly learn and understand excavator sensitivity and abilities, making the learning curve rather flat.
Xcavator first and early release utilizes the enormous and free Flickr/Yahoo public digital image library to perform its impressive capabilities. More advanced functionalities and tools for Xcavator will be released in the near future.
Xcavator is still limited in many aspects as it allows you to perform its magic only a certain pre-determined sets of images for now. Bear with it. This is just to give you an idea of what Xcavator will be able to do, while getting some positive comments and feedback from users like you on how exactly to make this tool most effective and easy to use for everyone else.
Xcavator is powered by award-winning CogniSign Intelligent Image Recognition Technology, designed for human-like image search and content management for large image databases.
Try for yourself Xcavator out, on this limited public demo: http://www.xcavator.net/index.html?sid=