Watson knows which one is the real killer whale fin. Do you?
Choose an image to find out:
A killer app - Watson Visual Recognition in action
Killer whale fins and the 'saddle patch' behind them are as unique as fingerprints - thanks to the infinite combinations of their size, shape and colour.
IBM Business Partner, TBSCG, has been working closely with a UK-based killer whale and dolphin charity to train Watson Visual Recognition to identify killer whales by their fins and markings.
The iDentEye service they developed enables charity members to upload killer whale images, as well as geo-location and other useful metadata, via a web app, to be processed by Watson.
The end result is a list of automatically recognized images that can be used by the charity for research, and for monetization through the extension of games, visual mapping and other services.
It's just one of the many innovative uses of Watson Visual Recognition.
How to enhance your own solutions with Watson Visual Recognition
Use our built-in capability
Watson’s ready-to-use Visual Recognition API has been pre-trained to recognise a huge range of scenes, objects, colours, food, text, content and other subjects that can give you insights into your own visual content. It can also detect faces in your images and provide age and gender estimates for each one.
Create your own classifiers
Want Watson to identify something new, like products, spare parts, rust, building damage…? Watson can be customized quickly, easily and cost-effectively for your own use case. In fact, you can start training your own custom model in minutes. With only a few images, Watson can learn any new object, person, or attribute. What’s more, the specific capability you create is entirely yours – it is not shared with anyone else.
It’s easy – and free – to get started
With our 30-day Lite Plan, you can make up to 7,500 API calls for free (up to 250 calls per day) and train a custom model using up to 5,000 images. Once the 30-day free trial is over and you upgrade to our Standard Plan it’s still incredible value, costing just £0.00121 per image tagging event!
Try Watson for yourself
Want to see just how good Watson is at classifying images? Load anything you like into the demo and see how Watson does.
From killer whales to broken boilers – Watson Visual Recognition can be used for virtually anything
Manufacturing, visual auditing, insurance, social listening, social commerce, retail, education – there are countless ways Watson Visual Recognition can be deployed to enhance your solutions.
Here are a few videos showing how Watson works and how easy it is to train.Watch how drones are used to film pylons, with the images analysed by Watson Visual Recognition to show if they need maintenance
See how easy it is to train Watson to spot rust on bridges
Training Watson to assess damage for insurance valuations
What could Watson Visual Recognition do for your customers?
Try the new With IBM portal
It’s designed to give ISVs of all sizes everything they need to embed IBM technology into their solutions, all in one place.
Visit With IBM today >
Helping ISVs on their journey to cloud based business growth...